Repository navigation
Agent pane: the live Thinking line starts on the icon column and stays the last row - #34370
Conversation
…g messages Fails on base: withWaitingRows does not exist and App.tsx appended the waiting-messages group after the live line. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-1f5b0310f68aec102ea5
…s the last row Leo: "the indent looks bad" / "thinking shimmer not being at the bottom looks bad". Drops the live line's text-column inset (#33157) and keeps it under the waiting-messages group and held guest prompts (withWaitingRows). Leo-ask: ask-f654b777b7 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-1f5b0310f68aec102ea5
|
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 Unclear: CI failed on
Failures:
Matched log linesNot re-run automatically: Written by red: web / Web status; a job it gathers is red on this change (its line here); do not merge until green; see lane-rules Shipped in:
|
|
Review (subagent claude-opus-5-5 medium): SHIP at 14e9fcb Checked row order, isLiveStep spacing, footer marks, Find and types; the first pass caught a readonly/mutable type error, fixed at this head. |
|
Found 44 test failures on Blacksmith runners: Failures
...and 34 more test failures. View all on Blacksmith
|
|
Review (codex gpt-6.1-sol xhigh, router-review): BLOCK at 14e9fcb BLOCK at 14e9fcb
These failures follow from the unchanged assertions; execution was unavailable because Bun/bunx are absent. Follow-ups:
|
…ew order and icon column Router review (gpt-6.1-sol) at 14e9fcb: the text-column and spacing browser tests still asserted the live line on the text column and over the waiting group. Leo-ask: ask-f654b777b7 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-1f5b0310f68aec102ea5
…from the group's last line Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-1f5b0310f68aec102ea5
…e kinds test does Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-1f5b0310f68aec102ea5
|
Review (codex, router-review): SHIP at 4cadcda Reviewer: codex gpt-6-astra high, fast tier. Follow-ups: 1. Each lands as an issue in this repo that links this receipt; the lane that opened the PR fixes it forward. SHIP at 4cadcda No correctness regressions found in static review. Tests could not run because Bun and dependencies are unavailable.
Follow-ups:
|
…w the live line kept last (base reds) (#34620) * gallery: message-row plays click the one-line row's first line (#34464) #34464 returned message rows to one line: the first line sits on the row after the name, and stacked (#33249) stays a gallery version. #34149's plays clicked .cv-agentmsg__below, the stacked first line, which the default no longer draws (agent-message-row#sent-opened, #sent-opened-long: no element), and message-line-look#opened waited for Copy under the opened message, which only the stacked version moves there (waitFor timed out). The plays click the first line's words on the row, never the recipient's chip, and wait for what the one-line look draws. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-3bab30503f0b9a7059e5 * gallery: shift rows follow the one-line message rows and the live line kept last incoming-message#minimal-fade-count, #minimal-inline-name and #minimal-open go back down to 0.0175, 0.0175 and 0.0624: #34149 raised them for #33249's stacked lines, and #34464's one-line rows measure 0.01749, 0.01749 and 0.0623 again on base (push run 41756372530). queued-agent-messages#subsession-outbox and composer#queued-from-agent-open open a waiting message above the running turn's live Thinking line, which #34370 keeps as the last row by design (Leo: "thinking shimmer not being at the bottom looks bad"), so opening it moves the live line down by what opened (0.0159, 0.0082). Their 0 rows were measured with the live line over the waiting group. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Agent-Session: https://cmux.com/deeplink/agent?id=aid-3bab30503f0b9a7059e5 --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…t spacing test's after() helper (#34577) * base red: the What's New brief's More line sits on the spacing scale (design tokens) check:design-tokens is red on feat since b7c9441 (#34477): the brief popover's .wn-pop-more used margin-top 2px, off the token scale. It now uses --ds-space-1 (4px), the same gap .wn-pop-all uses above it. * base red: transcript spacing test defines after() again (webviews typecheck) webviews typecheck is red on feat: test/transcript-spacing-kinds.test.ts:247 calls after(), which #34464 (1e8a45b) removed together with below() while #34370 (8557b10, written on a base that still had them) added a new use. Restore both helpers as #34370 measured with them; no other change. * base red: cloudRun test names a string in its template (vp lint restrict-template-expressions) #34592 (31ffec4) builds a call label with ${params.cwd}, an unknown; vp lint treats that as an error in the pre-push web lint, so the merge service's hook refused #34577's squash. String() it; no behavior change. * base red: format three webviews files the way vp fmt --check wants feat's emoji shortcode change (#34364) left App.tsx, emojiShortcodes.test.ts and composer-emoji.browser.test.ts unformatted, so vp fmt --check fails the pre-push web lint on every webviews push. Output of scripts/web-lint.sh --fix. * base red: the Cloud start source check accepts the formatted App.tsx call vp fmt puts cloudFolderGrant.granting(folder.cwd, pickFolder(...)) on several lines, and cloudRun.test.ts read it as one line, so it failed after the format commit. The pattern now allows whitespace between the arguments. bun test cloudRun.test.ts: 21 pass, 0 fail.
![Fix with [code]smith](https://pr-comments-assets.blacksmith.sh/codesmith/fix-with-codesmith-light-3x.png)
Summary
Leo, on the running turn's live line ("Thinking · 1m 23s total · 3.2K tokens" drawn under "Sent a message, finished 3 background tasks, ran 5 commands"):
Leo-ask: ask-f654b777b7
Reference: cmux-design-references
cmux/live/transcript-activity-summary-thinking-row(and a second capture of the live line drawn above a waiting message's "Send now" row).cv-worked.cv-worked--live, added in Agent chat: every status row's words on one column; session names read as words #33157). Its words now start on the icon column, the transcript's left edge, the way the shimmer sits in the Claude and ChatGPT apps, so it no longer reads like a second line of the row above it. The empty-slot markerdata-lead-slotand the gallery icon-column invariant's handling of it are gone too: the invariant now reads the line's words like any other text row.withWaitingRows(conversation/pendingMessages.ts) now puts them above a trailing live line (isLiveLine, model.ts). The line stays the last row, the same way a held prompt already sits above the line inside its turn.Risk: small and contained (CSS, one row-order helper, comments). Nothing changes for an ended turn. Spacing stays the same because the live line under the group's head is still "one more line of the run" (
isLiveStep).Design spec: design/chat-transcript.md ("One icon column": the live line now starts on the icon column).
Test
bun test src/agent-session/acpmux/conversation/pendingMessages.test.ts: the new test "the waiting messages draw over the running turn's live line" fails on base, wherewithWaitingRowsdoesn't exist and the group was appended after the line, and passes on the head (22 pass).Observed / Not run
scripts/web-lint.sh.Changelog
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
The live Thinking line now starts at the transcript's left edge and stays the last row, below waiting-message groups and held prompts. Its spacing is unchanged, and ended turns still place waiting messages at the end.
Written for commit 4cadcda. Summary will update on new commits.