`_renderMarkdown` wrote dir="auto" into its string, before the code pins and
the linkifier. An element carrying dir="auto" re-resolves its direction on
every insertion under it, and a block whose letters all sit in `dir` children
walks its whole subtree each time, so a code-only or path-only document went
quadratic, English ones included.
Measured in real Chromium through the real File Viewer, before and after:
9000 inline-code items 8.7 s to 0.35 s, 9000 bare paths 5.6 s to 0.23 s, a
209 KB out-of-workspace list 41.9 s to 1.4 s. All three now match master.
The pass is now `_applyBlockDirection()`, run last on the DOM by both
consumers. The direction results are unchanged. Tests pin that the rendered
string holds no dir="auto" and that the pass runs after the linkifier.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
An account sudoers lists by name, with a password, is routed to no-admin
because nothing can read sudoers without reporting to the administrator.
The route itself stays as chosen, but the reason no longer claims the
account may not use sudo, and every message that ends or sums up the route
names CODEMAN_NO_ADMIN=0 for that reason alone.
tmux and cloudflared are staged in the home folder, so a noexec /tmp no
longer reports a working build as one that does not run. USER gets a default
at the top of the script, which also covers the older lines this route can
reach. A trailing slash on CODEMAN_INSTALL_DIR no longer puts .new and .old
inside the install. The data directory prompt names a runtime kept one
prompt earlier.
The uninstall check reads the source and must never drive uninstall(): HOME
does not redirect systemctl --user, so a drive reaches the real service.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The summary and README said rerunning the script updates Codeman. It never
fetches, so that was a successful no-op. The summary, README, its zh-CN mirror
and the wiki now say git pull, then rerun, with the in-app route first.
Also corrects "printed once" (twice at a terminal, never off one), says a typed
answer beats a preset, adds the first-run setup, the 180 s wait, --no-wait and
the changeme refusal to Running-As-A-Service and Troubleshooting, and records in
architecture-invariants why docker/.env is generated FROM .env.example.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
With stdout captured (CI, provisioning, `| tee`) the log held the password twice
for a service that controls Docker on the host, while docker/.env, the stronger
copy, is 0600. Off a terminal the script now points at the file instead. The note
beside it no longer promises a second showing, which `--setup-only` never gave.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The refusals of $HOME, ~/.codeman and the checkout compare strings, so
`~/../me` or `<repo>/../repo/data` passed them: the first shares state with a
native install, the second lands in the build context, where `COPY . .` would
bake CLI logins into the image. Doubled slashes are collapsed and any `.` or
`..` segment is refused.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The first-run gate tested `! -f`, so a dangling symlink was silently replaced,
and a directory got a 0600 temp file holding a working password, a "Wrote"
message and exit 0, with setup repeating on every later run. Setup now runs only
when nothing is there. A symlink to a real file stays an existing install.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
In a single-quoted dotenv value a backslash escapes the closing quote, so a
password or data path ending in one passed validation and produced a docker/.env
Compose cannot parse. The error named an unrelated line, printed the password,
and a rerun never asked again. The env_quote comment claimed a single quote was
the only character such a value cannot hold, and the test stub's comment said
bash reads these values the way Compose does: both corrected, and a test now
reads generated values back by Compose's rules.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Compose prefers the shell environment over the env file, so with
CODEMAN_APPDATA_PATH or CODEMAN_PORT preset, a different typed answer went into
docker/.env and the summary while the stack ran on the preset: state split
across two folders and the next plain start mounted another home. Setup now
unsets everything it wrote, so the file it just printed is what Compose reads.
The test stub honours that precedence, and the questions get their first test
through a real pty.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
`docker info --format` exits 0 with the daemon unreachable on those clients,
so a user outside the docker group got a raw Compose error and a "network
hiccup" hint after docker/.env and the data folder had already been written.
`docker version --format` exits 1 on every release and keeps the same stderr
text, so the permission branch still matches. The test stub now answers `info`
the way those clients do.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The cases-folder refusal grew to 143 characters, and the status line is one
line with an ellipsis and no title, so at 513px on desktop it ended at "or
make the case with Ad..." and lost the instruction. .path-picker-status.error
now wraps: 2 lines on desktop, 4 on a phone, measured clear in Chromium at
both, and it covers the picker's other refusals too.
Tests: a pin on that rule (fails without the fix), and the option check now
reads each picker's whole options object instead of a fixed 400 characters,
which stopped short of the container-workdir picker's 829. Retitled to say
what it checks: Create New has the flag too.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
list-panes -t '=name' with no colon is a window target to tmux, so a
remote session that is gone resolved to a window of that name or to a
session the name is a prefix of, and the Edit history panel moved to that
session's conversation and stored its diffs. `=name:` is an exact session
match, as src/remote-pane-history.ts already uses. Measured on tmux 3.4.
The fake tmux in the test matched the name as a substring, so it passed
with or without the colon. It now needs the exact token and answers a bare
prefix the way real tmux does.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
With the folder field empty, Create New's Browse opens in the cases folder, and New
folder there answered "Use Add Case > New Case". There is no tab of that name (the
label is Create New), and a Create New user is already on it. The refusal now reads
right from Link Existing and from Create New's custom folder. Its test pin, the code
comment, the manual, api-reference, CLAUDE.md and architecture-invariants move with it.
The block() helper in the New folder picker test returned an empty string for a name
that is not in session-ui.js, so .not.toContain passed and a renamed picker dropped out
of the check. It now asserts the name was found.
The manual's path picker list said two places and left out Create New, which the New
folder paragraph names. It now lists every Browse that opens the picker, and that
paragraph is rewrapped to the page's width.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The setting told users the sandbox and approvals stay on. With it on, a
command in Codex's sandbox can call Codeman's API, which can start a session
outside the sandbox with no approval prompt. Measured on codex 0.162.0 with
`codex sandbox`: a touch the sandbox refused landed through POST /api/sessions,
and from inside a Codeman pane it landed with a password set too, because a
pane inherits CODEMAN_PASSWORD. Without that password the same call gets 401.
The settings row, the wiki, the skill and the changeset now say so, the wiki
Security page and security-architecture name it, and the docs scope the claims
to what the code does: app-wide, local launches only, not cron jobs, and
--no-daemon is a local launch's flag. Agent-CLIs gains the --no-daemon
trade-off. A test pins that no surface sells the sandbox as still in charge.
Text only: the switch, the clamp and the launch are unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The test waited for a real 1000 ms timeout and read the pid file afterwards.
On a loaded machine the kill could land before the fake ssh wrote it. It now
waits for the pid and sends the timeout's SIGTERM itself, which is all Node's
timeout does. Without exec the fake still survives and the test goes red.
The invariants said "the reads" go through buildRemoteEditHistoryCommand; the
image probes go through remoteProbePaths, on the same limiter.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The read script carries one case line per subagent cursor, and the whole
script is one argument of the ssh line, which Linux caps at 128 KiB. From
about 1000 subagents the command failed with E2BIG. The cursors have no cap
and the stored conversation is read first, so the session stayed unavailable
and not even a /clear got past it.
A read with more than 300 subagent requests now takes several round trips.
The cut is by name range, not only by count: the script takes a file no case
line names for a new one, so plain groups would read each other's files as
new. Each round trip walks only the names between its cuts, through one awk,
so a new file lands in exactly one range. Budget and meta budget carry over,
and a later round trip still lists its files, because a requested file missing
from the result loses its cursor. A failed round trip fails the whole read and
nothing of it reaches the store. A read of up to 300 runs no awk at all.
Each round of the 60-seed fuzz is also cut at a random size and compared with
the local reader. Run outside the suite too: bash --posix, BusyBox and mawk,
and bash 3.2 in a container, which is macOS sh. That shell cannot parse a case
inside $(...), so the names come from a shell function.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Applies the review of #766.
The live lookup ran tmux on the bare PATH of the ssh command. On a host
where tmux sits in ~/.local/bin or /opt/homebrew/bin it found no pane, the
2>/dev/null hid that, and the panel stayed on the launch conversation after
a /clear. It now goes through withRemoteTmuxPath(), inside the sh script and
scoped to the $(...), so the other tools keep the PATH they had. The fake
tmux of the test block moved to $HOME/.local/bin for the same reason: that
directory is first on the new PATH, so a real tmux is never asked.
The two ssh commands get exec. Node signals the local sh -c, and dash stays
the parent of a plain command, so ssh survived the timeout. A test with a
fake ssh shows the process dies now and survives without exec.
The comment at the meta budget claimed a meta past it comes with a later
read. A meta comes only with a subagent's first read, so it is never sent.
Docs: the store is named as the exception to "never copied to local disk",
remoteProbePaths and the remote transcript lookup are described, and the
line saying a remote session loads no store is gone.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The owner's answer to the PR's open question: the switch does not also pass
--sandbox workspace-write. Settings, wiki, skill and release note now state
the consequence: a folder Codex runs read-only keeps the network closed and
asks per command. A golden test pins the decision by name, and the stale
launch-defaults comments now name networkAccess.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Conflict in src/session-cli-registry-bridge.ts: kept the PR's gateVersion() (gates measured against the gated CLI's own version) and added master's Discord startup hint engine value.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A tcsh login shell worked on 1.48.0 only because the local sh rewrote
${SHELL:-/bin/sh} to this host's shell. With the launch verbatim, tcsh gets -i -l,
refuses it and every stock launch leaves a dead pane. The changeset and
docs/remote-sessions.md described that as a standing limitation; they now name it
as a change in this release, with what the override workaround costs.
Also: extraSshOptions joins the list of values now literal at launch, the stock pane
command is shown as /bin/sh -c in the docs and the builder's JSDoc, the invariants
state how the launch must reach the far end, the manual pages cover $ in an override,
and the invariants no longer say extraSshOptions refuses a bare $ (it never did).
Reworded the em-dashes on the invariants line this rewrites.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ownerClampedEnvKeys() read enabledClis(), so disabling a CLI dropped its
privileged keys from the clamp, and a revoked owner got a saved home back on
recovery or reboot restore. getCli() still returns a disabled entry, and a
session can outlive the switch.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
getCustomModelForPersist() trusted homeEnvVar to name a path, but the registry
types it as any env name, so a clis.json override could write a session's own
credential to state.json. Only an absolute path is written now.
A re-apply resolved the CLI's real home from what the session's map held, which
under a selection is the injected directory, and recovery resolved none at all.
An entry the own home lacks (settings.yaml) was then linked from the server's
home on an endpoint switch and on recovery. The route now reads the new
Session.ownEnvValue(), and recovery reads the clamped record through the same
bounded guard, under the bulk cap since it walks every saved session. Also
corrects the comment that said only a removed directory gets its links again.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
RemoteCommandOverridesSchema is .strict() and stopped at deepseek, so the API
refused a commands.omp or commands.copilot override although both modes can run
remotely and in a container (RemoteCommandMode, DockerCommandMode) and the launch
builders honour the key. The docs call the override the way out for a login shell
the stock launch cannot run, which was not true for those two modes.
The test walks the CLI registry, so the next CLI cannot miss the schema the same way.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The rename's checks inside mutateLinkedCases and the refused-registry
branches of rename and clone had no test: removing them left every
case route test green. A hook right after the folder move lands a link,
an unlink, a re-pointed entry or a torn file between the route's first
check and its registry write. Each new test fails when its check is
removed.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The lookup judges liveWorktree.path lexically and only when the pane moves,
while the route's probe and readdir both follow symlinks. A symlink put at
that path, or at a parent of it, without moving the pane re-rooted the listing
outside workingDir. Names and sizes leaked, while file-content and file-raw
still refused the paths.
The route now resolves both folders on every request and walks the worktree
only when it lies strictly inside the resolved workingDir, otherwise it walks
workingDir. Only the check uses the resolved paths: the walk keeps the session's
spelling, so a workspace reached through a symlink still gets paths relative
to workingDir.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
transcriptPageKeys applied to every pi, and a pi in `tuiMode: regular` binds
PageUp/PageDown to its editor. It emits no mouse enable, so the flag stays
false, and on a fresh pane the buffer is hollow: the wheel sent PageUp, the
caret of a multi-line draft moved, and the next keystroke landed on the wrong
line. Measured on real pi 1.1.0, 1.4.0 and 1.5.0. Regular mode is not exotic,
it only takes one line in pi's settings. Telling the two modes apart would
need a new capability, since the paging gate takes no id, so pi has no
fallback: a fullscreen pi with "Wheel scrolls local history" on stays where
master left it.
altScreen, the half of the fix that does the work, was pinned by nothing:
reverting it left 1328 tests green. The strip decision table reads its
expected value from the same registry entry. pi now has its own block using
the bytes it really emits, which also pins that focus reporting survives.
The fullscreen default dates from pi 1.0.0, not 1.4.0, so the comments say
that, and the sites still calling pi narrow-strip or main-screen by default
are corrected. Where a comment listed CLIs it now names the capability.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Handing the launch to bash verbatim exposed `"${SHELL:-/bin/sh}"` to the far
end for the first time. tmux runs a one-argument command through the remote
default-shell, not /bin/sh as the comments said, and fish refuses `${...}`:
on a fish host, or a bash login with `set -g default-shell fish`, every new
pane died with status 127. master only survived because the local sh had
already rewritten that token to this host's shell path.
Codeman's own text now goes to `/bin/sh -c` as separate tmux arguments, which
tmux execs without a shell. A commands.* override stays one argument and keeps
the default-shell it was written for: under `/bin/sh` a `source` or `[[` in an
override dies with 127 on bash and zsh hosts.
The tests read the argv tmux would receive through a real sh, so they need no
fish: every stock launch must reach /bin/sh and an override must not. The
docker half of bashCArg was pinned by nothing, so it gets its own test.
Comments that described the removed double-quote layer are corrected, and
remote-hosts.ts no longer claims overrides are seeded from the default text
or that they rescue nushell and elvish.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude Code's EnterWorktree moves the running process into
<repo>/.claude/worktrees/<name>, while Session.workingDir stays on the main
checkout. The Files panel kept walking workingDir, so it listed a tree the
agent had left.
paneCwd already follows the move, measured on Claude Code 2.1.318 in both
directions, so nothing new has to track the agent. It also follows every cd
in a shell, so a new SessionState.liveWorktree narrows it to a linked
worktree, read from the .git file without starting git.
Only a folder strictly below workingDir is ever an answer, and the listing
only starts there: returned paths stay relative to workingDir, so the
preview, raw and edit routes are unchanged and confinement does not widen.
A worktree outside the session folder stays unfollowed.
The panel asks with root=live, shows a strip naming the worktree with a
switch to the session folder, and follows the agent in and out while open.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
pi 1.4.0 is an alt-screen TUI that enables mouse tracking once at start.
Codeman's replay (tab switch, reload) drops that mode in xterm
(mouseTrackingMode none, baseY 0), so the wheel hit an empty local buffer
and stayed dead until a resize made pi redraw.
Same shape as copilot: altScreen strip-mux-and-mouse (the server records
the stripped DECSETs as cliMouseTracking, which survives the replay),
wheelForward always, transcriptPageKeys (PageUp/PageDown fallback for a
stale flag, measured on pi 1.4.0). _shouldForwardWheelToApp treats pi like
copilot. Registry pins and docs updated.
Tests: 2 new in terminal-scroll-routing (red before). Full run: 12989
passed, 5 red from the de_DE locale / XDG env (custom-model-run-menu-ui,
git-worktrees, edit-history-sync), all green under LANG=C.UTF-8.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
(cherry picked from commit 53e4136d9f4401e9250ac6a122599ffcb2a57e8a)
respawn-pane runs `bash -c "<ssh …>"` through the local sh, so every `$` in
the remote command is expanded on the Codeman host. withRemoteTmuxPath
(`export PATH="$HOME/.local/bin:…:$PATH"; tmux …`, 5b9445de) therefore sends
the Codeman host's home and PATH to the remote host: a remote user with a
different home (tmux in ~/.nix-profile/bin) gets "tmux: command not found" and
every new remote pane exits 127. `${SHELL:-/bin/sh}` in the remote launch was
the local shell for the same reason.
Remote and docker commands are already quoted for the far end, so they now go
to `bash -c` single-quoted (bashCArg); local commands keep the double-quoted
form. Two tests run the captured respawn-pane line (create and respawn)
through a real sh with a foreign HOME/PATH and check that ssh gets `$HOME` and
`$PATH` unexpanded.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 2818e9455489b9acd65841258df4cd035609954c)
Codeman itself always lived under $HOME; administrator rights were only
spent on what it needs first (system packages for Node.js, tmux and git,
Homebrew on a Mac). An account without those rights could not install it:
the installer died in ensure_sudo, or in Homebrew's own installer.
install.sh now has a no-admin route, taken by itself when the account has
no administrator rights (no sudo binary; outside sudo/wheel/admin and
`sudo -n true` fails; a refused sudo password; a Mac account outside the
admin group with no writable Homebrew) and forced with --no-admin or
CODEMAN_NO_ADMIN=1:
- Node.js: the official tarball from nodejs.org, checked against the
SHASUMS256.txt published beside it, into ~/.codeman/runtime/node.
- tmux: the tmux project's own build (tmux/tmux-builds 3.7c), checked
against a SHA-256 pinned in the installer, into ~/.codeman/runtime/bin.
- Both are linked from ~/.local/bin (never over a file that is not ours),
and the runtime is on PATH and on the LaunchAgent's PATH.
- git is never installed. Without a working git the source comes as an
archive, the install is marked with .install-source, and it updates by
re-running the installer: the new tree is built beside the old one and
swapped in only after the build succeeded. A later re-run on a machine
that has gained git turns it back into a clone.
- Nothing on this route runs sudo, Homebrew or `xcode-select --install`,
and the route is never decided by `sudo -v`.
What it cannot supply is checked before the first download: a C/C++
toolchain on Linux (node-pty 1.1.0 compiles there), musl, and macOS below
15 (the tmux build's deployment target).
Along the way:
- check_git no longer trusts `command -v git`. On a Mac without the Xcode
Command Line Tools /usr/bin/git is a stub that opens Apple's install
dialog; it is judged by `xcode-select -p` and never executed. Any other
git is asked for its version.
- scripts/git-hooks.mjs answers "not a checkout" without running git, for
the same reason: postinstall calls it on every npm install.
- The in-app updater knows the install kind `archive` and points such an
install at the installer instead of `npm i -g aicodeman@latest`.
- download() and download_to_stdout() resolve DOWNLOADER themselves. Only
main() used to set it, so `install.sh cloudflared` died on `set -u` with
the message swallowed by the step log.
- The AI CLI menu leaves out a Homebrew-only install command where there
is no Homebrew.
Verified with the real installer: a bare Ubuntu 24.04 container as an
unprivileged user (install, a PTY round trip through the API, update by
re-run, cloudflared, uninstall), the same without a compiler (stops before
any download), with a broken then working git, with sudo present but not
usable, and on macOS 26 (arm64) in an isolated HOME with sudo, brew, git,
the compilers, launchctl and tailscale replaced by tripwires. Not verified
on a real standard macOS account: that account was an administrator.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Applying a custom model writes the endpoint's values over the keys it
injects, and clearing the model unset those keys. A session that had set
one of them itself lost that value: its own DSH_HOME, CODEX_HOME,
GROK_HOME or GEMINI_API_KEY was gone once the model was cleared, and so
was CLAUDE_CONFIG_DIR, which silently moved a Claude session onto the
default account. Switching endpoints did the same.
setCustomModel() now records what the session held under each injected
key and restores it on clear and on switch. Quick-start passes the
caller's values to the Session constructor so they are seen before the
selection lands.
The selection overwrites the only other copy, tmux's setenv, so the
session's own home directory has to survive a Codeman restart on its
own. __customModel therefore stores that one path, the single exception
to persisting keys only, and recovery runs it through
clampEnvOverridesForOwner() so a revoked grant cannot bring it back. Any
other own value, a credential included, stays in memory only.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The probe-stall test and the removal test were written against the path-only deps. With the session argument the first failed outright and the second passed without reaching the probe. Both now use the real shape; removing the isRemoved() guard turns the second red. The route comments no longer call remote sessions unsupported.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Review round 3 (all low):
- the "written last" test left solo newer than conv-1 for the rest of the
file, so the live test passed by mtime alone. solo's mtime is reset in a
finally, and the live test now runs with the live transcript older than
the stored one (a /resume into an older conversation before it writes).
- the server wiring guard matches the remote branch of the ternary as one
expression, so swapped branches fail it.
- note on whole-second `-nt` (BusyBox, bash 3.2) at the lookup.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Review round 2: with the stored conversation tried before the launch id, a
respawn that resumed the launch conversation after a /clear was missed once
claude exited again: the cleared-to conversation still had a file and won,
so the edits made after the resume were never read. No fixed order of the
ids is right both after a /clear and after such a resume. Without a live
conversation the lookup now takes the most recently written transcript of
all known ids (stored, claudeSessionId, id); the live one still goes first.
Also from the review:
- the remote no-transcript note and the docs name login profiles too
(.profile, .bash_profile): sshd's `bash -c` reads neither, like .zshrc;
- a wiring guard pins that server.ts looks a remote transcript up with
remoteTranscriptIds and leaves the local lookup on the hook path;
- the browser tests undo the injected remote session in a finally.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Review findings on the remote edit history:
- A forward read may carry 13 MiB and "Load older" 29 MiB; the flat 30 s
read timeout cut those off below ~1 MiB/s. The timeout now adds the
transfer time at 256 KiB/s (remoteEditSourcesTimeoutMs). It ends a hung
read, it does not cut a slow one.
- The transcript lookup looked only under $HOME/.claude. It now uses an
absolute CLAUDE_CONFIG_DIR of the ssh command's environment, else
$HOME/.claude, for both the projects and the live sessions/<pid>.json.
A value set only in .zshrc stays invisible (the lookup is a plain sh);
documented, and the panel's remote no-transcript note says so instead
of promising the next prompt (a remote host sends no hooks).
- After a remote /clear claudeSessionId stays the launch id, so once the
remote claude exited the lookup fell back to the cleared conversation
and the panel flipped back. The sync now hands the lookup the stored
conversation, tried right after the live one (remoteTranscriptIds).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Codeman maintainer