Repository navigation
test(transcript): Claude cohesion variant measures its turn's Worked for - #35410
Conversation
#34774 folded a finished Claude turn under one Worked for, but transcript-cohesion still asserted the Claude variant draws no header, which left the file red on feat-cmux-next. The Claude variant now checks the same 22px header and 16px reply gap as closed Codex. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-82f54cd5a2280880c43b
GitHub Actions
Inactive
|
cmux status (one comment, edited in place; each line is updated by the job that owns it) Review App of this head (app.zip, kept 1 day)
Gallery before/after for
States
Pick other states with a Passes: CI passes on CI passes on Written by web and size checks: green Shipped in:
|
|
Owned by an agent session · session |
GitHub Actions
Inactive
|
Review (codex, router-review): SHIP at eb6204c Reviewer: codex gpt-6.1-sol xhigh, fast tier. Follow-ups: 1. They land as checklist items in this PR's one follow-up tracking issue in this repo ( SHIP at eb6204c Visual verdict: no images; Leo tries the build
Follow-ups: F1. |
|
Finding outcome: comment-6631604013/F1 dismissed test-only change; the rendered UI is unchanged and screenshots never block (the fleet run's measured geometry is in the body). |
|
Found 3 test failures on Blacksmith runners: Failures
|
![Fix with [code]smith](https://pr-comments-assets.blacksmith.sh/codesmith/fix-with-codesmith-light-3x.png)
![cmux-bot[bot]](https://avatars.githubusercontent.com/u/171392238?s=60&v=4){kind=small}
GitHub Actions
Inactive
GitHub Actions
Inactive
A newer verdict blocks 443045e (no independent SHIP at the head).
|
Review (codex, router-review): SHIP at 443045e Reviewer: codex gpt-6-astra high, fast tier. Follow-ups: 1. They land as checklist items in this PR's one follow-up tracking issue in this repo ( SHIP at 443045e Visual verdict: no images; Leo tries the build
Follow-ups:
|
GitHub Actions
Inactive
GitHub Actions
Inactive
A newer verdict blocks 8707f1d (no independent SHIP at the head).
|
Review (codex, router-review): SHIP at 8707f1d Reviewer: codex gpt-6-astra high, fast tier. Follow-ups: 1. They land as checklist items in this PR's one follow-up tracking issue in this repo ( SHIP at 8707f1d Follow-ups:
|
|
For awareness: #35537 makes a finished Claude turn draw no whole-turn Worked for by default again, which follows leo-workspace#3287 (Leo: "you can't do it for Claude ... it's just too lossy if you do it for Claude"). The fold #35113/#34774 gave Claude is now an opt-in, the |
|
Finding outcome: comment-6633355342/F1 dismissed test-only change with no rendered UI change; the fleet run's measured geometry is in the body. |
…ked for again (undoes #35410's expectation) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-92fa78a91e29868c5eaa
…its version, plus the 480 s file limit Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-fb75b7af9902231028aa
…-turn Worked for is an opt-in (#35537) * Agent pane: a finished Claude turn keeps its runs in place; its whole-turn Worked for is an opt-in Leo (2026-11-04): "you can't do it for Claude ... it's just too lossy if you do it for Claude". The claude row of TURN_FOLD_STYLES stops folding a whole turn; the Worked for fold #35113 made Claude's default is the opt-in arm of a new claude-turn-fold experiment. The tests and gallery plays #34774 moved to Claude's Worked for go back to its per-run lines. Codex, Gemini and Muse are unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-92fa78a91e29868c5eaa * gallery: the per-harness fold entry covers the rows it draws Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-92fa78a91e29868c5eaa * test: a finished Claude turn of the send loop draws its run in place, no Worked for Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-92fa78a91e29868c5eaa * test(transcript): the Claude cohesion variant draws no whole-turn Worked for again (undoes #35410's expectation) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-92fa78a91e29868c5eaa --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Fixes the base red in
webviews/test/transcript-cohesion.test.tson feat-cmux-next (web / react-apps-base-check (2)reported "webview test red step (base): test/transcript-cohesion.test.ts", which then charged unrelated PRs such as #34881 with "claude keeps the activity rhythm (light)").Regressing PR: #34774 ("a finished Claude turn folds under one Worked for, its runs inside"). It gave a finished Claude turn the same Worked for header Codex has, per Leo: "ofc claude code doesn't actually have the worked for... but i think we should be able to do that replacing the top level thingy maybe". #34607 had written the cohesion test while Claude still drew no header, so its
claudevariant assertedheaders.count() === 0. The two landed about an hour apart and neither ran against the other, which left bothclaude keeps the activity rhythmtests failing on base (2 of 48 failed, reproduced at 6e9fc2c).Fix: the test is stale against the settled design, so the test changes and the product code doesn't. The
claudevariant now goes through the same checks ascodex-closed: a 22px header, a 16px gap to the reply, and 38.375 centre to centre (step-to-prosein design/transcript-spacing.md). The gallery note and header comment drop "no whole-turn Worked for". No spacing value changes.Fleet run at this head: 48 pass, 0 fail (claude light/dark measured
headerHeight 22, replyGap 16, replyCenterGap 38).Agent session: agent session
🤖 Generated with Claude Code
Open in [code]smith · Need help on this PR? Tag
@codesmith-botwith what you need.Summary by cubic
Fixes the stale transcript-cohesion test so the
claudevariant measures its turn'sWorked forheader instead of asserting none exists. PR #34774 folded finished Claude turns under the same header Codex has, which left this test red onfeat-cmux-next.claudevariant to check the same 22px header and 16px reply gap as closed Codex.Written for commit 8707f1d. Summary will update on new commits.