fix(orchestrator): Keep lifecycle rows visible in the work log by mwolson · Pull Request #4762 · pingdotgg/t3code

mwolson · 2026-07-28T17:37:04Z

Summary

Keep important lifecycle outcomes visible in the web timeline:

Running and stopped tool rows survive work-log filtering and compact views.
A terminal provider error remains visible when a failed turn produces no
assistant message.

These are provider-agnostic presentation fixes. In each case, the server
projection was correct and the web timeline hid the only useful row.

Problem and Fix

Problem and Why it Happened	Fix
Work-log filtering treated explicit `inProgress` and `stopped` lifecycle rows like rows with no lifecycle signal.	Keep neutral rows when the provider reported an explicit running or stopped lifecycle, while continuing to hide signal-less noise.
Running and stopped rows had no accurate indicator treatment because they previously never reached the row component.	Show a spinner only for an executing tool. Keep stopped tools visible without giving them a success indicator.
Compact work logs could drop active or stopped rows when newer ordinary entries consumed the visible-row budget.	Pin executing and stopped rows, then fill the remaining budget with the latest ordinary rows.
Settled-turn folding retained a terminal assistant message, but a provider failure can end with only a failed error work item. The error then appeared only in the sidebar.	Exclude failed provider-error items from the fold's hidden-entry set. Recovered retry rows remain ordinary foldable work.

Defensive Fixes

Problem and Why it Happened	Fix
The CTM test fixture supplied a fallible `ServerSettings` layer where the provider-session test runtime requires an infallible layer.	Convert that test-only layer with `Layer.orDie`. Production runtime behavior is unchanged.

Validation

vp check: passed
Full workspace and mobile typechecks: passed
session-logic.test.ts, MessagesTimeline.logic.test.ts, and
MessagesTimeline.test.tsx: 105 tests passed after the terminal-error append
Independent GPT-5.6-Sol high review of the terminal-error change: no findings
Isolated renderer test with a real Codex turn:
- A foreground command appeared with a running spinner.
- Stop interrupted the turn.
- The command row remained visible without a spinner or success check.
A live failed provider turn confirmed that retries and persistence were
correct while settled-turn folding hid the final capacity error.

Notes for review

A stopped row intentionally has no lifecycle glyph. The component already uses
XIcon for failure, while the red stop marker belongs to the separate interrupt
lifecycle row. Inventing another glyph would turn a visibility fix into a design
change.

The terminal-error exception is narrow: it requires an error work item with a
failed lifecycle. A recovered provider-retry row remains foldable.

The original neutral-row behavior arrived in #3022 ui/messages-timeline. Mobile
does not share the affected web filtering or settled-turn folding paths.

Note

Low Risk
UI-only timeline presentation and tests. No auth, data, or orchestration behavior changes beyond a tiny test-layer .pipe(Layer.orDie) tweak.

Overview
Keeps in-flight and interrupted tool rows visible in the chat work log, and stops inferring their status from the whole turn.

Neutral rows with a reported inProgress or stopped lifecycle now render; only rows with no lifecycle signal stay hidden. Collapsed groups pin executing and stopped tools so a later completed row cannot bury them. Pending/waiting tools stay unpinned and show no glyph (the old minus/"Empty" indicator is gone).

Row indicators now follow the item’s own status: a spinner only when structuredPayload.status === "running", no success check for stopped tools, and aria-label suffixes for running/stopped/failed. activeTurnInProgress is removed so a settled turn cannot mark live background work as complete.

Failed provider-error work entries also stay unfolded when there is no assistant message.

^{Reviewed by Cursor Bugbot for commit 38109f8. Bugbot is set up for automated code reviews on this repo. Configure here.}

Note

Keep in-flight and stopped tool rows visible in collapsed work log groups

Adds selectCollapsedWorkLogEntries in MessagesTimeline.logic.ts to pin executing and stopped entries while showing only the most recent ordinary entries within MAX_VISIBLE_WORK_LOG_ENTRIES
Adds lifecycle helpers workEntryHasReportedLifecycle, workEntryIsExecuting, and workEntryShouldRenderInWorkLog in session-logic.ts to distinguish genuinely running tools from neutral rows with no lifecycle signal
Reworks SimpleWorkEntryRow in MessagesTimeline.tsx to show a spinner only for running tools, a checkmark for success, an X for failure, and no icon for pending/waiting; stopped state is conveyed via accessible name suffix
Removes the activeTurnInProgress prop from MessagesTimeline and ChatView; lifecycle indicators are now driven solely by entry data
deriveTurnFolds keeps terminal provider error entries visible in settled turns instead of folding them away
Risk: workEntryShouldRenderInWorkLog now hides neutral work log rows that lack any lifecycle signal — any caller expecting those rows to appear in collapsed groups will no longer see them

^{Macroscope summarized 38109f8.}

coderabbitai · 2026-07-28T17:37:12Z

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: d63aaf22-b302-4873-a096-e3875d635abe

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

Create PR with unit tests

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands.}

Keeping these rows in the work log exposed that the row component had no treatment for them, so they borrowed the wrong one. A stopped tool rendered a green check with a "Completed" tooltip once the turn settled, because the settled-turn fallback treats any neutral row as success. A running tool rendered the "Empty" dash, which is meant for rows with no content. Both now render their own state: a spinner tooltipped "Running" while the turn is live, and a square tooltipped "Stopped" for an interrupted tool. An `inProgress` row that outlives its turn is also shown as stopped rather than completed, since the turn ended without the tool ever reporting a result. Reported by Macroscope on pingdotgg#4762, which caught the stopped-as-completed half.

Codex review of pingdotgg#4762 caught that inferring "stopped" from a settled turn is wrong. `activeTurnInProgress` is thread-wide, and background work legitimately outlives the root run: `RunExecutionService` explicitly keeps ingesting a late background command completion after root terminal, in the order running -> root-finalized -> completed. The inference would have labelled that live work "Stopped" and then jumped it to "Completed". Both indicators now read the item's own lifecycle and nothing else. A genuine interrupt already terminalizes the item at the projection layer, which is what made the inference look unnecessary rather than harmful. Also from that review: - The work group's filter is now `workEntryShouldRenderInWorkLog`, a production helper, so the predicate tests exercise the shipped policy instead of a copy of the expression that could drift from it. - The lifecycle joins the row's accessible name ("Ran command …, running"), since the indicator sits inside the row button and did not contribute to it. The spinner is decorative rather than a nested role="status" announcing a generic "Loading". - Component tests cover running during an active turn, running after settle, stopped in both states, and a signal-less row staying hidden. Tooltip text is absent from static markup, so they assert on the icon and accessible name, as the existing failed-indicator test does.

Keeping these rows in the work log exposed that the row component had no treatment for them, so they borrowed the wrong one. A stopped tool rendered a green check with a "Completed" tooltip once the turn settled, because the settled-turn fallback treats any neutral row as success. A running tool rendered the "Empty" dash, which is meant for rows with no content. Both now render their own state: a spinner tooltipped "Running" while the turn is live, and a square tooltipped "Stopped" for an interrupted tool. An `inProgress` row that outlives its turn is also shown as stopped rather than completed, since the turn ended without the tool ever reporting a result. Reported by Macroscope on pingdotgg#4762, which caught the stopped-as-completed half.

Codex review of pingdotgg#4762 caught that inferring "stopped" from a settled turn is wrong. `activeTurnInProgress` is thread-wide, and background work legitimately outlives the root run: `RunExecutionService` explicitly keeps ingesting a late background command completion after root terminal, in the order running -> root-finalized -> completed. The inference would have labelled that live work "Stopped" and then jumped it to "Completed". Both indicators now read the item's own lifecycle and nothing else. A genuine interrupt already terminalizes the item at the projection layer, which is what made the inference look unnecessary rather than harmful. Also from that review: - The work group's filter is now `workEntryShouldRenderInWorkLog`, a production helper, so the predicate tests exercise the shipped policy instead of a copy of the expression that could drift from it. - The lifecycle joins the row's accessible name ("Ran command …, running"), since the indicator sits inside the row button and did not contribute to it. The spinner is decorative rather than a nested role="status" announcing a generic "Loading". - Component tests cover running during an active turn, running after settle, stopped in both states, and a signal-less row staying hidden. Tooltip text is absent from static markup, so they assert on the icon and accessible name, as the existing failed-indicator test does.

…eachable Codex round 2 on pingdotgg#4762. Once `workEntryShouldRenderInWorkLog` admits neutral rows only when they carry a reported lifecycle, the row's neutral handling can never fire: `showNeutralIndicator` and the settled-turn success fallback both require the opposite. So `turnSettled`, `hasReportedLifecycle`, `showNeutralIndicator`, the "Empty" dash branch, and the compound success expression were all dead. Success is now just `workEntryIndicatesToolSuccess`, and the row no longer reads thread-wide turn state at all. Also from that round: - Tests pin "no stopped indicator" properly. A stopped row could have regressed to the dash while every assertion still passed; there is now a negative `lucide-minus` check, and the running case asserts the spinner carries neither `role="status"` nor `aria-label="Loading"`, so the accessibility fix cannot silently revert. - The comment claiming both states get their own indicator is gone, as is the overstatement that a "Run interrupted" divider always sits beneath a stopped row: `cancelled` maps to stopped too and has no such row.

…eachable Codex round 2 on pingdotgg#4762. Once `workEntryShouldRenderInWorkLog` admits neutral rows only when they carry a reported lifecycle, the row's neutral handling can never fire: `showNeutralIndicator` and the settled-turn success fallback both require the opposite. So `turnSettled`, `hasReportedLifecycle`, `showNeutralIndicator`, the "Empty" dash branch, and the compound success expression were all dead. Success is now just `workEntryIndicatesToolSuccess`, and the row no longer reads thread-wide turn state at all. The running test also asserts the spinner carries neither `role="status"` nor `aria-label="Loading"`, so the accessibility fix cannot silently revert. The comment claiming both states get their own indicator is gone, as is the overstatement that a "Run interrupted" divider always sits beneath a stopped row: `cancelled` maps to stopped too and has no such row.

macroscopeapp · 2026-07-28T18:59:22Z

Approvability

Verdict: Approved at 38109f8

Macroscope's review found this PR approvable — UI bug fix with well-scoped changes to work log display logic. Adds helper functions and smarter collapsing to keep running/stopped tools visible. Includes comprehensive test coverage. No backend or API changes.

^{You can add or adjust custom eligibility rules. Learn more.}

Macroscope on pingdotgg#4762. `projectedWorkEntryStatus` collapses `pending`, `running`, and `waiting` into `inProgress`, so keying the spinner off `toolLifecycleStatus` made a queued tool and an approval request blocked on the user both claim to be running, with a "running" tooltip and accessible suffix to match. That was a regression from this PR: before it, those rows reached the indicator chain with no branch that applied and rendered nothing. `workEntryIsExecuting` reads the projected item's real status instead, so only a genuinely running tool spins. `pending` and `waiting` rows still render, since showing in-flight work is the point of the PR, but they claim nothing. The test helper now derives the projected status from the collapsed lifecycle value, so a fixture cannot assert on a pairing the projection never emits; the previous version let a `stopped` row carry a `running` item.

Dismissing prior approval to re-evaluate 7d5e55c

…eachable Codex round 2 on pingdotgg#4762. Once `workEntryShouldRenderInWorkLog` admits neutral rows only when they carry a reported lifecycle, the row's neutral handling can never fire: `showNeutralIndicator` and the settled-turn success fallback both require the opposite. So `turnSettled`, `hasReportedLifecycle`, `showNeutralIndicator`, the "Empty" dash branch, and the compound success expression were all dead. Success is now just `workEntryIndicatesToolSuccess`, and the row no longer reads thread-wide turn state at all. The running test also asserts the spinner carries neither `role="status"` nor `aria-label="Loading"`, so the accessibility fix cannot silently revert. The comment claiming both states get their own indicator is gone, as is the overstatement that a "Run interrupted" divider always sits beneath a stopped row: `cancelled` maps to stopped too and has no such row.

Macroscope on pingdotgg#4762. `projectedWorkEntryStatus` collapses `pending`, `running`, and `waiting` into `inProgress`, so keying the spinner off `toolLifecycleStatus` made a queued tool and an approval request blocked on the user both claim to be running, with a "running" tooltip and accessible suffix to match. That was a regression from this PR: before it, those rows reached the indicator chain with no branch that applied and rendered nothing. `workEntryIsExecuting` reads the projected item's real status instead, so only a genuinely running tool spins. `pending` and `waiting` rows still render, since showing in-flight work is the point of the PR, but they claim nothing. The test helper now derives the projected status from the collapsed lifecycle value, so a fixture cannot assert on a pairing the projection never emits; the previous version let a `stopped` row carry a `running` item.

cursor

Cursor Bugbot has reviewed your changes using high effort and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

Want fixes drafted automatically? Bugbot Autofix can create code changes for findings. A team admin can enable Autofix in the Cursor dashboard.

^{Reviewed by Cursor Bugbot for commit 8b32d74. Configure here.}

macroscopeapp

One issue found: the removal of the activeTurnInProgress prop leaves a stale call site in this file's own test, which no longer typechecks. Everything else in the diff (the Spinner running indicator, the , running/, stopped accessible-name suffixes, and the pinned collapsed work-group selection) is consistent with the existing indicator branches in SimpleWorkEntryRow and the shared UI primitives.

^{Posted via Macroscope — UI Consistency}

Dismissing prior approval to re-evaluate dd1f712

CI Check fails on current CTM because layerTest now surfaces ServerSettingsError while the ProviderSessionManager test layer requires never. orDie matches the other test layers.

Dismissing prior approval to re-evaluate 19d3389

Dismissing prior approval to re-evaluate 38109f8

github-actions Bot added size:S 10-29 changed lines (additions + deletions). vouch:unvouched PR author is not yet trusted in the VOUCHED list. labels Jul 28, 2026

macroscopeapp Bot reviewed Jul 28, 2026

View reviewed changes

Comment thread apps/web/src/components/chat/MessagesTimeline.tsx

mwolson changed the title ~~fix(web): Keep in-flight and stopped tool rows in the work log~~ fix(orchestrator): Keep in-flight and stopped tool rows in the work log Jul 28, 2026

github-actions Bot added size:M 30-99 changed lines (additions + deletions). and removed size:S 10-29 changed lines (additions + deletions). labels Jul 28, 2026

mwolson force-pushed the fix/web-timeline-inflight-tool-rows branch from caf6f38 to d30b4a7 Compare July 28, 2026 17:55

mwolson force-pushed the fix/web-timeline-inflight-tool-rows branch from 5a7eae3 to 257f810 Compare July 28, 2026 18:36

mwolson force-pushed the fix/web-timeline-inflight-tool-rows branch from 257f810 to 365949a Compare July 28, 2026 18:55

mwolson marked this pull request as ready for review July 28, 2026 18:55

macroscopeapp Bot reviewed Jul 28, 2026

View reviewed changes

Comment thread apps/web/src/components/chat/MessagesTimeline.tsx Outdated

macroscopeapp Bot previously approved these changes Jul 28, 2026

View reviewed changes

github-actions Bot added size:L 100-499 changed lines (additions + deletions). and removed size:M 30-99 changed lines (additions + deletions). labels Jul 28, 2026

macroscopeapp Bot previously approved these changes Jul 28, 2026

View reviewed changes

mwolson mentioned this pull request Jul 28, 2026

fix(orchestrator): Keep active web timeline visible while work is queued #4800

Closed

4 tasks

macroscopeapp Bot previously approved these changes Aug 5, 2026

View reviewed changes

juliusmarminge force-pushed the t3code/codex-turn-mapping branch from 25de21d to 0af2a6e Compare August 7, 2026 12:10

mwolson force-pushed the fix/web-timeline-inflight-tool-rows branch from 4b122f7 to 8b32d74 Compare August 8, 2026 18:41

cursor Bot reviewed Aug 8, 2026

View reviewed changes

Comment thread apps/web/src/components/chat/MessagesTimeline.logic.ts

juliusmarminge force-pushed the t3code/codex-turn-mapping branch from 22bd872 to a27c1cc Compare August 10, 2026 17:05

mwolson force-pushed the fix/web-timeline-inflight-tool-rows branch from 8b32d74 to b387e85 Compare August 11, 2026 13:25

juliusmarminge force-pushed the t3code/codex-turn-mapping branch 2 times, most recently from a186d64 to 5b1a115 Compare August 12, 2026 23:19

mwolson mentioned this pull request Aug 13, 2026

fix(orchestrator): Close and drain OpenCode 1.x tool items #4786

Closed

juliusmarminge force-pushed the t3code/codex-turn-mapping branch 4 times, most recently from 519c42a to 4c55679 Compare August 17, 2026 10:28

mwolson force-pushed the fix/web-timeline-inflight-tool-rows branch from b387e85 to ede1305 Compare August 19, 2026 15:37

macroscopeapp Bot reviewed Aug 19, 2026

View reviewed changes

Comment thread apps/web/src/components/chat/MessagesTimeline.test.tsx

macroscopeapp Bot previously approved these changes Aug 19, 2026

View reviewed changes

mwolson added 4 commits August 19, 2026 14:04

fix(server): Make ServerSettings test layer infallible

d0ab308

CI Check fails on current CTM because layerTest now surfaces ServerSettingsError while the ProviderSessionManager test layer requires never. orDie matches the other test layers.

fix(orchestrator): Keep in-flight and stopped tool rows

f8fbb98

fix(orchestrator): Keep active rows visible in compact work logs

e03e92e

fix(web): Drop leftover activeTurnInProgress timeline test prop

19d3389

mwolson force-pushed the fix/web-timeline-inflight-tool-rows branch from dd1f712 to 19d3389 Compare August 19, 2026 18:05

macroscopeapp Bot previously approved these changes Aug 19, 2026

View reviewed changes

fix(orchestrator): Keep terminal provider errors visible

38109f8

mwolson changed the title ~~fix(orchestrator): Keep in-flight and stopped tool rows in the work log~~ fix(orchestrator): Keep lifecycle rows visible in the work log Aug 20, 2026

macroscopeapp Bot approved these changes Aug 20, 2026

View reviewed changes

t3dotgg added vouch:trusted PR author is trusted by repo permissions or the VOUCHED list. and removed vouch:unvouched PR author is not yet trusted in the VOUCHED list. labels Aug 24, 2026

Navigation Menu

Uh oh!

fix(orchestrator): Keep lifecycle rows visible in the work log - #4762

fix(orchestrator): Keep lifecycle rows visible in the work log#4762
mwolson wants to merge 5 commits into
pingdotgg:t3code/codex-turn-mappingfrom
mwolson:fix/web-timeline-inflight-tool-rows

mwolson commented Jul 28, 2026 •

edited

Loading

Uh oh!

coderabbitai Bot commented Jul 28, 2026 •

edited

Loading

Review skipped

Uh oh!

Uh oh!

Uh oh!

macroscopeapp Bot commented Jul 28, 2026 •

edited

Loading

Uh oh!

cursor Bot left a comment

Uh oh!

Uh oh!

macroscopeapp Bot left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Conversation

mwolson commented Jul 28, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Problem and Fix

Defensive Fixes

Validation

Notes for review

Keep in-flight and stopped tool rows visible in collapsed work log groups

Uh oh!

coderabbitai Bot commented Jul 28, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review skipped

Uh oh!

Uh oh!

Uh oh!

macroscopeapp Bot commented Jul 28, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Approvability

Uh oh!

cursor Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

macroscopeapp Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

mwolson commented Jul 28, 2026 •

edited

Loading

coderabbitai Bot commented Jul 28, 2026 •

edited

Loading

macroscopeapp Bot commented Jul 28, 2026 •

edited

Loading