From review round 8: "Mission complete!" with confetti and cheering
helpers sat over balloons saying the page will not work as opened, a
helper's tests have not run, or a problem is left. The final panel is now
calm (no confetti, a neutral robot) whenever work was still out at the
turn's end, nothing worked, or Claude's last words name an open problem
(`page.openNote`: a key point about something wrong, not done, not
checked or not fixed); the last of these reads "Done, with a catch".
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Two agents read only the comic of the eleven recorded sessions (one who
has never written code, one engineering manager skimming), wrote down
what they believed, then compared with the transcripts. Both were misled
in the same places, and the cause was the same: the final panel kept
Claude's first sentence or two, and what a reader must not lose sat
further into its last message.
- Claude's last words now keep key sentences from anywhere in its final
message (read in full, past the label's clip): an open problem or a
quirk, something not yet done, checked or confirmed, work still out,
what you asked for or expected, a correction of an earlier warning
(with the sentence it refers to). A bare "I didn't" counts only when
nothing stronger is there; table rows, list intros and questions never.
"I built the pomodoro timer. Opening the page: it won't work if you
double-click index.html…"; "…That note makes the current behaviour
clear, but it doesn't fix it."; "As you expected, git describe --tags
then failed because the repo has no tags."
- A turn that ended with helpers or a background command still out says
so: "Done for now", and "Claude's part is done in 36 seconds. Its 2
helpers finished later and reported back." (`finishedLater`).
- A failure you asked for, or that Claude said it expected, reads as the
plan: "Test first: the first run fails as planned, the second
passes.", "The tests don't even load, as planned." A loop that failed
once on a test and once on another check says both, and the commits
made inside a loop are part of its story; the final panel counts the
turn's commits ("5 commits, uploaded").
- One or two files read are named ("Claude checks README.md." in a
turn a report started); a tool lookup reads "Claude gets the tools it
needs ready." again (it was overwritten) without its query; the
advisor is "a stronger AI model", and one that answered with an error
"gets no answer"; a commit message loses its `feat:` prefix in the
plain caption.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
- The bar on a phone took four rows (216 px of a 664 px screen) with the
close button alone on the last; it now takes three (170 px): title and
close, the turn, the view and Save image. The drawer starts exactly
under it and follows it when the phone turns (the bar is observed).
- A panel below the fold had no buttons until it was scrolled to (only
drawn panels get click regions), so Tab skipped it. A quiet frame now
draws one such panel, a band at a time; its entrance still plays when
it comes into view.
- "Fix and test: the first run fails, the second passes." was chosen by
rounds, so a loop with three runs (a failing test, then ruff, then a
pass) said it under ✗ ✗ ✓. It now needs exactly two runs; that loop
reads "Fix and test, 2 rounds: 2 runs fail, then it passes.", and the
corpus pins every loop's marks.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
From review round 5 (a phone with touch):
- Touch screens: every control at least 44 px (prev/next, close, the
Story/Director's cut toggle, Save image, the final panel's chips, drawer
buttons and links), and the small type at 13 px.
- The drawer on a phone starts below the bar (it cut through it), closes by
dragging its head down, and keeps its head small in landscape.
- The turn picker says where you are ("Turn 2 of 5"), learns about new
turns while you look at an older one, and a turn you picked yourself is
not announced as news.
- A check whose command also wrote files offers "See the 2 files it wrote".
- Test counts are read from the whole output: a later command's lines had
pushed pytest's "5 passed" out of the kept tail.
- A save names its commit: "Claude saves the work: "feat: add wrap() for
modular wrapping into a range"." (from -m, -am, --message or the
here-document of git commit -F -; a trailer is never the subject).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Review round 5 had an agent read only the comic of the nine real sessions
(20 turns), write down what it believed happened, and only then compare
with the transcripts. Three panels told it something false, one hid the
failure the user asked to see:
- A push that went through, followed by a check that failed in the same
command, drew as a failed push; the output (`[new branch]`, `->`) now
decides: "Claude saves the work in 2 commits and uploads it, but a step
after that fails."
- A formatter's "would be reformatted" preview, printed before it fixed
the file, made a clean run a failure: the last verdict wins.
- A blocked command ended in confetti and "Mission complete!": a run panel
where everything failed says "Claude tries a command, and it is
blocked.", and a turn where nothing worked ends "Turn over" / "Nothing
Claude tried worked, so the job is not done." with a neutral robot.
- The loop stamped only the final verdict (POW ×2), so a first run that
failed on purpose never showed. It now marks every test run red or green,
and says it: "Fix and test: the first run fails, the second passes." /
"Build and test, 4 rounds. A check fails once, then everything passes."
Also:
- Files written inside a check or a save reach the plain caption ("Claude
writes 4 files; the tests pass.").
- Balloons keep the next sentence when the first reports a problem ("…left a
stray file. I'll remove it…") or is very short ("Not done. The harness
blocked sleep 65…"); Claude's last words keep a caveat that follows them
("Nothing is committed because this folder isn't a git repository.").
- The advisor never shares a panel with helpers that do work; ToolSearch
reads "Claude gets the tools it needs ready."; Codeman's own case file is
not drawn as a file Claude read; looking around shows what Claude meant
(the call's description), not raw commands; a montage at the start of a
turn reads "Claude gets started".
- Words: boards count tests ("19 tests"), "uploads it" replaces "sends it
up", and the prompt balloon keeps 400 characters.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Stepping through 40 turns made 46 long tasks of 51 to 286 ms, enough for
the app's guard to drop the terminal's WebGL renderer for the rest of the
page's life. Drawing switched off made none: the cost was canvas raster at
paint time, which the per-frame JS budget never saw. Now:
- Panel backgrounds are shared between panels and turns (kind × size ×
palette), and a new one is built in four bands, one per frame (a wide
halftone at 2x was ~35 ms in one go; a band is ~8 ms).
- At most two panels are drawn per frame, the first one served rotating so
none starves; clocks still run for all.
- Measured on 20 turns stepped 40 times at 2x: 1 long task (52 ms), against
39; the settled overlay is still ~2% of a core.
Also from round 4:
- The turn picker says when older turns were not loaded (`older` in the
response: the transcript is longer than the part that was read).
- A compaction after the last turn closes that turn; a local command's
output is the final panel's balloon ("Compacted"), without its markdown
ticks, and a turn under a second reads "Done.".
- Exact durations keep their seconds past an hour, so the two captions never
disagree on the minute ("1 h 10 min." / "1 h 9 min 39 s").
- A turn that arrives already finished is announced as new too.
- Whole files written by here-documents read "Claude writes 3 files."; the
advisor's caption reads "a second opinion" instead of repeating its name.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Nine transcripts from the review rounds, each a real Claude Code worker
building a small program (a Python CLI with a fix loop, a web page with
node:test, a bash script with a background helper, a long turn with a
prompt added mid-turn, teammates then an interrupt, a question then tests
first then a push, a text plan with a scratch dry run, /compact then plan
mode, a test failed on purpose). Each is cut down to the lines and fields
the reducer reads, with every path, user name and email rewritten and the
file contents of Reads left out; replaying them before and after gives
the same pages.
The expected file beside each pins every page (panel kinds, captions,
balloons, counts), so a change that alters a story fails with a diff of
exactly what changed; STORY_CORPUS_UPDATE=1 regenerates them. The beats
the reviews were about are also stated outright.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Codeman maintainer