Skip to content

feat(dashboard): say when a conductor mode run is done or stopped - #37058

Merged
bolichen97 merged 1 commit into
mainfrom
feat/conductor-mode-run-done-state
Nov 14, 2026
Merged

bolichen97 merged 1 commit into
mainfrom
feat/conductor-mode-run-done-state

Conversation

@bolichen97

@bolichen97 bolichen97 commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor

Problem / Motivation

Goal: A person looking at a conductor mode chat can tell when its run is done, stopped or still working.

Today the Board never says a run is finished. After every milestone is accepted it still reads "Done when ...", which looks pending. A rejected or abandoned milestone shows the green done check and counts toward "X of Y". A run that stopped with milestones open looks stalled or stuck. The Sessions row says "Done" when every worker session finished, without reading the work ledger, so a run stopped with open work can read "Done".

Why it matters

Conductor mode runs a goal through many worker sessions over hours. If the person cannot see that the run ended, and whether it ended with everything accepted or with work left, they either keep waiting on a finished run or miss work that was dropped.

Not a goal

  • Saying who stopped a run: a conductor's own stop and a person's stop record the same outcome on the loop.
  • A new word on the Sessions row, a notification, or a chat-header badge.
  • Any change to the conductor's prompt, tools or the work ledger's records.

What changed (motivation → approach → change)

A conductor already ends its run by sending one final report and then stopping its patrol loop. So the gateway can tell a finished run from facts it already holds, with nothing new stored:

Board has items, and the conductor's loop is... Every item closed Some item open
active still going still going
not active done stopped
  • dashboard/conductor_mode_run.py works this out as run (settled, open, accepted, not_done, at_limit, patrol_on). GET /api/crew-board carries it, and keeps a new goal with no item yet unsettled through the header's goal_items_used. The root row's crew_work carries it once its board has an item; the row cannot see the goal's start, which the spec lists as a known limit.
  • The Board's Overview opens with one line: Paused, Waiting for you, Working, Done or Stopped (lib/conductorModeRun.ts puts what is live first). An unsettled board reads Working only while the patrol is on, so a new goal waiting for its go-ahead with no milestone yet draws no line. Done and Stopped add one plain sentence with no numbers. The line adds no control; the panel's Back row stays the one way back. No line claims a final report exists, because a stop at a limit or by a person sends none. Done with milestones left keeps a muted check; Stopped shows a stop sign.
  • A rejected or abandoned milestone reads "Not done" with a muted ✕ and is not counted as done.
  • The Sessions row reads "Done" only when run.settled is done; a root with no ledger keeps the every-worker-done rule.
  • Twelve new strings in every catalog under website/src/i18n (English in locales/en.manual.json), the run field on the board and slot types (types/index.chat.ts, api/crewBoard.ts), a capture page for the five states, and the rule written down in the crew-mode spec and docs/work-ledger.md.

The alternative, a "finished" record the conductor writes, needed a ledger field and prompt text in four places and could go stale after a new goal, so it was not taken.

Backwards compatibility

Compatible: run is a new optional field on two reads; an older dashboard ignores it, and this dashboard keeps today's behaviour when it is absent.

Removes nothing: no field, key, record or path is removed. The row's "Done" and the milestone check now follow the ledger, which corrects what they claimed.

Tests

  • test/test_conductor_mode_run.py: the verdict table, a new goal with no item yet staying unsettled on the board read, the loop reading (active, stopped, budget stop, no loop), patrol_on off for a new goal waiting on its first item, the row's run with the fold's forgotten closes, no run for an empty board, and the board read's run.
  • website/src/lib/conductorModeRun.test.ts: the state order.
  • website/src/pages/chat/command-center/ConductorModeCrewPanel.run.test.tsx: the Overview line in each state, Not done rows, no second way back beside the panel's Back row, one notice per failed reopen, Waiting for you while the main chat holds its own tool approval, no claim of a final report (a stop at a limit included), no line without a goal.
  • website/src/test/ChatSidebar.conductorLane.test.tsx: Done follows run.settled; a stopped run keeps the usual line.
  • ConductorModeCrewPanel.test.ts: an abandoned or rejected item is not_done.

The new run-line and row tests fail on the base commit, except the no-goal test, which pins today's behaviour. The backend tests passed 20 times at -n0, 10 times at -n 4 beside test_crew_work_summary.py and test_crew_board_asks.py, and once in a shuffled order.

The final head is a content-identical rebase of the head both local review lanes passed (db5581004f): git diff <base>..<head> | git patch-id --stable gives 394a03a56ea8fe59a951fef974e0235fc138a8f5 for both.

Manual verification

Rendered the real Board panel from website/capture/conductor-mode-run-states.html in Chromium, before (base) and after, for each state, and looked at every image.

Screenshots / video

Before → after, Board Status view at 360 px.

State Before After
Done before done after done
Done, some not done before done-some after done-some
Stopped before stopped after stopped
Working, Waiting for you and Paused
State Before After
Working before working after working
Waiting for you before waiting after waiting
Paused before paused after paused

Related Issues

Closes #36912

Checklist

  • At most two commits (one is the norm), with a Conventional Commits title (feat|fix|docs|style|refactor|perf|test|chore|ci|build|revert: ...)
  • Existing tests pass and new tests added for new functionality
  • Self-review completed; code follows project style guidelines
  • Documentation updated (if applicable)
  • No secrets, credentials, or internal references in the diff

@bolichen97
bolichen97 requested review from a team and KiroCrewCaptain as code owners November 14, 2026 03:45
@bolichen97
bolichen97 requested review from cruisercohen and removed request for a team November 14, 2026 03:45
@github-actions github-actions Bot added the readiness: checking Automated validation is still running label Nov 14, 2026
@github-actions

github-actions Bot commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor

Design Review (Opus 5.5) — ✅ PASS

Design-level review of 47d3e1f9a1de19784d843e9af294e49a591099ac — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS does not.

Design-Verdict: PASS

Nothing blocks: the verdict is computed on each read from facts the gateway already holds, nothing new is stored, and the field is additive.

Notes

  • The row's verdict covers the whole board, not just the current goal. So when a new goal starts, the Sessions row keeps showing the previous goal's "Done" until the new goal's first item lands. Once the work fold records when a goal starts, scope the row's counts to that goal to close this limit; the spec and code already note it as a shortcut.

[DESIGN-REVIEWED] 47d3e1f

Description read

[DESCRIPTION-READ] 7f5cf1eda2bca467a9d7dc1cca03c9cb53e6d156662c28050c4977831fee83d4

That sha256 names everything this verdict read: the PR description, plus the 1 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with .github/scripts/pr-description-capture.sh over the same inputs; if it differs, the description or that evidence changed after this verdict was formed and any finding drawn from either is unproven.

@github-actions

github-actions Bot commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor

First Principles Review (Opus 5.5) — ✅ PASS

Premise-level review of 47d3e1f9a1de19784d843e9af294e49a591099ac — why this exists and whether the shipped surface is the smallest honest version. Updated in place on each push. A BLOCK verdict blocks PR readiness; PASS does not.

First-Principles-Verdict: PASS

Nothing blocks; the run field ships two counts no surface reads.

Notes

  • Item 5 — zero consumers: run_facts returns "open": open_items, "accepted": accepted, but nothing outside tests reads run.open or run.accepted (grepped run\??\.(open|accepted) in website/src and src: 0 hits; settled, not_done, at_limit and patrol_on each have at least one reader), and work-ledger.md now documents both as public fields. Subtraction: delete open and accepted from run_facts, ConductorModeRun, the capture fixtures and the two doc lists, and keep settled, not_done, at_limit and patrol_on.

What this change ships

Inventory (7 items) — 6 justified

Intent: a person watching a conductor mode chat can tell whether its run is done, stopped or still working. This is an ADDITION, and it also fixes two places where the dashboard claimed more than was true (#36912).

  1. The Board's Overview opens with one line: Paused, Waiting for you, Working, Done or Stopped — justified
  2. Done and Stopped each add one plain sentence, with a separate wording for a stop at a time or budget limit — justified
  3. A rejected or abandoned milestone reads "Not done" with a muted ✕ instead of the green done check — justified
  4. The Sessions row says "Done" only when the gateway's run verdict says done — justified
  5. A new run field on GET /api/crew-board and on the root row's crew_work — zero consumers for open/accepted
  6. Twelve new strings in every locale catalog — justified
  7. A capture page that renders the Board in each of the five states — justified

[FIRST-PRINCIPLES-REVIEWED] 47d3e1f

Description read

[DESCRIPTION-READ] 400b5c363e368173161e08903de7c3095abd2db0ae0ccfe426ae1e8fa0745ce9

That sha256 names the PR description this verdict read. Recompute it with .github/scripts/pr-description-capture.sh; if it differs, the description changed after this verdict was formed and any finding drawn from the description is unproven.

@github-actions

github-actions Bot commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor

UX Review (Opus 5.5) — ✅ PASS

UX-level review of 47d3e1f9a1de19784d843e9af294e49a591099ac — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS does not.

UX-Verdict: PASS

Nothing blocks. The blind reader read all five run lines and both "Not done" rows the way the diff means them, and every state appears in a screenshot.

Notes

  • Done with some milestones not done: the blind reader said the bold "Done" next to "Some milestones were not done" is "easy to misread shot-04 as a success". Drop the extra "Finished." from conductor_mode_run_done_some so it reads "Some milestones were not done." On its own that does not fix the bold "Done" sitting beside "not done": the only difference from a full success is the check's colour.
  • Stopped vs Paused: the reader could not tell them apart. Paused "tells me nothing" and has "no resume button". Stopped's "Write to the main chat" is not a link. Give Paused a one-line next step to match Stopped, and say where the main chat is (the Back row).
  • No screenshot shows the conductor_mode_run_stopped_limit sentence ("It stopped at a time or budget limit…") or the Sessions row "Done" now following run.settled. Add both to the capture page.

[UX-REVIEWED] 47d3e1f

Description read

[DESCRIPTION-READ] bbe48022bcb64dd3a0660d6f4eb1e4cc5313f21cec59df7e25f90076efe9531c

That sha256 names everything this verdict read: the PR description, plus the 15 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with .github/scripts/pr-description-capture.sh over the same inputs; if it differs, the description or that evidence changed after this verdict was formed and any finding drawn from either is unproven.

@github-actions

github-actions Bot commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor

GPT 6.1 Review — ✅ no blocking findings

GPT 6.1 completed its review of 47d3e1f9a1de19784d843e9af294e49a591099ac and found no blocking issues.

This comment is updated in place on each push.

Review details

No findings.
[GPT-REVIEWED] 47d3e1f

False positive or not applicable? A repository writer can comment:
/ai-review override gpt 47d3e1f9a1de19784d843e9af294e49a591099ac false-positive: <reason, 20 to 500 characters>

@github-actions

github-actions Bot commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor

Opus 5.5 Review — ✅ no blocking findings

Reviewed 47d3e1f9a1de19784d843e9af294e49a591099ac — this comment is updated in place on each push.

Review details

No findings.

[OPUS-REVIEWED] 47d3e1f

Verdict parsed from the review's SHA-scoped output markers for commit 47d3e1f9a1de19784d843e9af294e49a591099ac.

False positive or not applicable? A repository writer can comment:
/ai-review override fable 47d3e1f9a1de19784d843e9af294e49a591099ac false-positive: <reason, 20 to 500 characters>

@github-actions github-actions Bot added readiness: action required A blocking check or review needs attention and removed readiness: checking Automated validation is still running labels Nov 14, 2026
Base automatically changed from refactor/rename-ultra-to-orchestrator-mode to main November 14, 2026 04:30
@github-actions

github-actions Bot commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor

Sensitive Change Review

  • No sensitive path changed; nothing to check.

@github-actions github-actions Bot added the merge conflict Branch has merge conflicts with its base — author must resolve before merge label Nov 14, 2026
@bolichen97
bolichen97 enabled auto-merge (squash) November 14, 2026 04:50
@bolichen97
bolichen97 force-pushed the feat/conductor-mode-run-done-state branch from 493b884 to 265776b Compare November 14, 2026 05:33
@bolichen97

bolichen97 commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor Author
  • A green "Done" contradicts a run that left work undone span=bc56b64eeda0

fixed in 265776b: a Done run with any rejected or abandoned milestone now draws a muted check, never the green one; the green check stays only for "Every milestone was accepted."
The same class was swept over every run-line and milestone icon: no other green mark sits beside a sentence that says work was left.
Pinned by ConductorModeCrewPanel.run.test.tsx (the done-some case asserts the icon is muted and not crew-tone-ok); the after-done-some screenshot in the description shows it.

@bolichen97

bolichen97 commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor Author
  • The Stopped icon looks like a control span=3cad48204e10

fixed in 265776b: Stopped now draws a stop sign (CircleStop) instead of the hollow square that read as a checkbox.
The other run-line glyphs were checked for the same class (pause bars, warning triangle, spinner, check) and none is a box; the not-done milestone badge is the same filled badge as the existing quiet and stuck ones.
Pinned by ConductorModeCrewPanel.run.test.tsx (the stopped case asserts lucide-circle-stop). The Paused state is now in the screenshots too.

@github-actions github-actions Bot removed the merge conflict Branch has merge conflicts with its base — author must resolve before merge label Nov 14, 2026
@KiroCrewCaptain

KiroCrewCaptain commented Nov 14, 2026 •

Copy link
Copy Markdown
Collaborator

Code change approver recommendation: A person should decide

Why: the decider estimates a 12% chance that this change introduces a defect, within the 12% limit for approving on its own, but it changes 970 lines, more than the 400 it may approve on its own.

Additive read-only run verdict computed from the ledger's item states plus the conductor's loop state, sent on GET /api/crew-board and on the root row's crew_work. Checked against the code: closed counts include the fold's forgotten closes and follow the existing state lifecycle (remember, retain, test reset). A loop start or stop already refreshes the conductor's own board through nudge_wiring → worker_loop_moved. Legacy headers skip the goal_items check. Rejected and abandoned milestones become not_done, stay among the open rows and leave the done count. The row's Done needs run.settled==='done' and still yields to anything running or needing the person. Every behaviour change has a test. Nothing blocks.

An automated recommendation, not a review: a maintainer still decides. It approves on its own only when the estimated chance of a defect is at most 12% and the change has at most 400 changed lines. Decider v3c (217fe607) · commit 47d3e1f9a1 · 970 changed lines. If this recommendation is wrong, reply here: replies are how the approver is evaluated.

@bolichen97
bolichen97 force-pushed the feat/conductor-mode-run-done-state branch from 265776b to 3409902 Compare November 14, 2026 06:29
@bolichen97

bolichen97 commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor Author
  • A new goal inherits the previous run's completion span=5914f3109104

fixed in 3409902: the board read passes the ledger header's goal_items_used to run_facts, so a new goal that has created no item yet stays unsettled (Working, or no line) instead of reading the old goal's Done; a header from before that count is not gated.
The Sessions row builds run from the work fold, which holds no goal start and no goal item count, so it cannot apply the same gate without a new crew-log fold field; that is out of this PR's scope and is recorded as a known limit in the crew-mode spec row and the shortcut: comment in run_facts.
Pinned by test/test_conductor_mode_run.py (test_the_board_read_carries_the_run records a new goal and asserts settled == "" until its first item closes); red on 265776b, green on 3409902.

@bolichen97

bolichen97 commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor Author
  • Budget-stopped runs assert that an unwritten final report exists span=7f14b9820b0c

fixed in 3409902: the sentence that said the final report is in the main chat is removed from every run state and its key from every catalog, because a stop at the time or budget limit sends no report and neither does a stop by a person, and the board cannot tell those apart.
The run line keeps only its plain state sentence and the Back to main chat link; no remaining string claims a report exists.
Pinned by ConductorModeCrewPanel.run.test.tsx (the at-limit, stopped and way-back cases assert no "final report" text); red on 265776b, green on 3409902.

@bolichen97
bolichen97 force-pushed the feat/conductor-mode-run-done-state branch 2 times, most recently from ff3d1ab to db80d1a Compare November 14, 2026 08:33
@bolichen97

bolichen97 commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor Author
  • Second Back to main chat control span=058bddd6e07a

fixed in db80d1a: the run line no longer draws a control of its own; the panel's Back row (CrewPanelBack) stays the one way back to the main chat, as its doc says.
The separate useCrewBack instance and its reopen notice that only that button used are removed with it.
Pinned by ConductorModeCrewPanel.run.test.tsx (with a crew member on screen there is exactly one crew-panel-back and no button inside the run line).

@bolichen97
bolichen97 force-pushed the feat/conductor-mode-run-done-state branch from db80d1a to 606bbaf Compare November 14, 2026 09:37
@bolichen97

Copy link
Copy Markdown
Contributor Author

Backend Tests (3.12, 2) on this PR fails test/tracing/test_tracing_otlp.py (test_a_traced_turn_reaches_a_loopback_receiver_with_its_vault_header, test_the_exit_flush_sends_spans_still_queued). This PR touches no tracing code; the same tests are red on main since #37005 and are tracked in #37169. Waiting for that fix to land on main, then this PR rebases onto it.

@bolichen97
bolichen97 force-pushed the feat/conductor-mode-run-done-state branch from 606bbaf to 58e13e3 Compare November 14, 2026 12:40
@github-actions github-actions Bot added readiness: checking Automated validation is still running and removed readiness: action required A blocking check or review needs attention labels Nov 14, 2026
@bolichen97
bolichen97 force-pushed the feat/conductor-mode-run-done-state branch from 58e13e3 to 261c176 Compare November 14, 2026 13:19
@bolichen97

Copy link
Copy Markdown
Contributor Author
  • A new goal waiting for its go-ahead reads Working span=278a25701d6a

fixed in 261c176: run now carries patrol_on, and conductorModeRunState reads an unsettled board with milestones as Working only while the patrol is on; a new goal with no item, no patrol and nothing running draws no line.
The class is every reader of an unsettled run: the Board's run line was the only one that inferred "patrol on" from it (the Sessions row reads only settled === 'done'). Pinned by conductorModeRun.test.ts, ConductorModeCrewPanel.run.test.tsx and the board-read case in test/test_conductor_mode_run.py.

@github-actions github-actions Bot added readiness: passed Eligible automated validation passed for the current revision and removed readiness: checking Automated validation is still running labels Nov 14, 2026
@bolichen97
bolichen97 force-pushed the feat/conductor-mode-run-done-state branch from 261c176 to a5ecb46 Compare November 14, 2026 14:35
@github-actions github-actions Bot added readiness: checking Automated validation is still running and removed readiness: passed Eligible automated validation passed for the current revision labels Nov 14, 2026
@bolichen97

Copy link
Copy Markdown
Contributor Author

Two reds on a5ecb4644c, neither in this PR's files:

The previous head 261c176186 (same patch-id) had every check green; this head is a content-identical rebase taken because main changed the i18n catalogs this PR also edits.

@github-actions github-actions Bot added readiness: action required A blocking check or review needs attention and removed readiness: checking Automated validation is still running labels Nov 14, 2026
The gateway derives a settled verdict from the work ledger item states
and whether a loop is still active on the conductor session: done when
every item is closed, stopped when some are open. GET /api/crew-board
and the root row crew_work carry it as run. The Board Overview opens
with one run line, a rejected or abandoned milestone reads Not done,
and the Sessions row says Done only when the run is done.

Closes #36912
@bolichen97
bolichen97 force-pushed the feat/conductor-mode-run-done-state branch from a5ecb46 to 47d3e1f Compare November 14, 2026 18:56
@github-actions github-actions Bot added readiness: checking Automated validation is still running readiness: action required A blocking check or review needs attention readiness: passed Eligible automated validation passed for the current revision and removed readiness: action required A blocking check or review needs attention readiness: checking Automated validation is still running labels Nov 14, 2026

@KiroCrewCaptain KiroCrewCaptain left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Approved. On head 47d3e1f the review read the whole diff and checked the callers of each changed function. It confirmed that patrol_of only runs on the event loop and that a missing nudge service reads as off. It also checked that the new _closed cache is cleared beside every _items write, and that the take-away changes to milestoneState and the Sessions row allDone have no readers outside this diff. In a scratch copy the backend tests passed 24 of 24, and the three related files passed 59 of 59 three times in a row. A mutation run showed that test_a_closed_count_change_alone_still_moves_the_row fails when the change check in crew_work.remember is removed.
Done well: "done" comes from facts the gateway already holds, so it stores nothing that can go stale. A rejected or abandoned milestone also no longer shows the green check.

  • Optional: no test covers the legacy header branch in work_ledger_board.py:801. Replacing it with goal_items=record.goal_items_used fails 0 of 113 tests. One board read whose conductor.json has no goal_items_base key would pin it.
  • Optional: nothing outside tests reads run.open or run.accepted. You could drop them, or give the First-Principles note a disposition that says why they stay.
  • Optional: patrol_of counts any active loop on the slot, but the sibling conductor_looping counts only the work-ledger patrol. Use one definition, or say in work-ledger.md that any watch keeps the run Working.

@bolichen97
bolichen97 merged commit 89c02b1 into main Nov 14, 2026
154 of 156 checks passed
@bolichen97
bolichen97 deleted the feat/conductor-mode-run-done-state branch November 14, 2026 21:23
@github-actions github-actions Bot removed the readiness: passed Eligible automated validation passed for the current revision label Nov 14, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Say when a Conductor Mode run is done or stopped

2 participants