Skip to content

fix(crew): draw the conductor's drawer board from its work ledger - #18957

Closed
iamwhatever wants to merge 2 commits into
mainfrom
fix/conductor-panel-fold-board
Closed

iamwhatever wants to merge 2 commits into
mainfrom
fix/conductor-panel-fold-board

Conversation

@iamwhatever

@iamwhatever iamwhatever commented Oct 10, 2026 •

Copy link
Copy Markdown
Contributor

Problem / Motivation

Goal: The goal conductor's side-panel task board stays current on its own, without the conductor having to remember to republish it.

The crew page's side panel (Dashboard preview off) shows only what the conductor last sent with panel_publish. When a patrol cycle skips the publish, the board goes stale for hours: new items never show up and finished ones keep their old state. The work ledger already holds every item the whole time.

Why it matters

The board is where a person looks to see what needs them. A stale board hides new work and new questions until the person asks the conductor in chat. Asking the agent to "remember to update" does not fix that, because forgetting is the failure.

Not a goal

What changed (motivation → approach → change)

Part Before After
Task list, state, PR last publish work ledger, every read
Step of a queued task hand-written none lit until a worker starts it
Needs you last publish items with question / blocked
Done count, updated last publish work ledger
next line last publish last publish; muted Last plan (<time>) once the ledger moved on
Crew with a ledger, no publish empty state board from the ledger
Open drawer re-read panel_published only conductor drawer only: also a 60 s foreground poll

The board was hand-written each cycle, so a skipped publish was a stale board. The drawer read (dashboard/handlers/agent_panel.py, _panel_record) already rebuilds the pipeline conductor's board from the work fold. It now does the same for kirocrew-conductor through the new provider build_conductor_board in conductor_board_contract.py. Published data is read only for next and repo. The ledger board is built only when the slot names this crew alone and a configured crew of that exact name owns the slug (otherwise the board is served as published), and its worker-written text goes through the same redaction a publish gets (src/kiro_crew/agent_panel.py: scrub_data, derived_record).

The template agent_panel_templates/kirocrew-conductor.html shows an old next muted, as Last plan (<next_from>). A queued item has an empty step, so its bar lights nothing. The drawer (pages/members/CrewWebview.tsx) re-reads the conductor's board on a 60 s foreground poll, because a ledger entry sends no frame that names this query; the poll keys on the record's template or the conductor's slug (pages/members/dashboardRefetch.ts) and reuses the Dynamic Dashboard tab's DASHBOARD_FALLBACK_REFETCH_MS. Every other drawer stays push-only. The goal-conductor skill (builtin_skills/goal-conductor/SKILL.md) no longer asks for a republish each cycle. Docs updated in docs/feature-map/README.md.

Backwards compatibility

Compatible: panel_publish accepts the same data; a published board is still served when no work ledger exists, and every other template is served exactly as published.

Removes nothing: the published task list is no longer the source of the conductor board, but it is still stored and read; only the fields the ledger owns are replaced on read.

Tests

  • test/test_conductor_board_from_fold.py (new): an item created in the ledger appears on the drawer with no panel_publish; items created after the last publish appear without another; needs-you comes from a worker's question; a stale next gets next_from; a board with no ledger is served as published; nothing is synthesized without a ledger or on a shared slot.
  • Red-before: on origin/main the three ledger-to-drawer tests fail (assert None is not None, the old one-item list); all six pass here.
  • test/test_agent_panel_routes.py: the single-read counter stub forwards the new keyword.
  • test_every_item_the_fold_kept_is_on_the_board: the task list and the done count cover the same items.
  • test_a_published_board_on_a_shared_slot_never_takes_the_ledgers_tasks and test_a_ledger_board_is_served_only_to_the_crew_that_owns_the_slug: another crew's ledger never reaches this crew's drawer, and a caller-chosen member name gets an empty panel.

Manual verification

Rendered the real drawer document through _panel_record against a real crew log (one publish, then three ledger items, one asking a question) and captured it at 420 px, before and after.

Screenshots / video

Before (main): the drawer shows only the last publish, one task, nothing needs you.

before

After (this head): three tasks from the ledger, U's question in Needs you, 1 of 3 done, the old next muted as Last plan (…), and queued V with no step lit.

after

Related Issues

Closes #18956

Refs #18919

Pattern harvest

Rule candidate: review-prompt
Pattern: a UI surface whose state an agent must re-send each cycle, while a log fold already holds that state.

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

Contribution License Agreement

Contributions are accepted under the project's Apache-2.0 license.

@iamwhatever
iamwhatever requested review from a team and KiroCrewCaptain as code owners October 10, 2026 22:16
@github-actions github-actions Bot added the readiness: checking Automated validation is still running label Oct 10, 2026
@github-actions

github-actions Bot commented Oct 10, 2026 •

Copy link
Copy Markdown
Contributor

First Principles Review (Opus 5.5) — 🟡 CONCERNS

Premise-level review of 45c576d3f804ba7f0cc83f4b63be26cefda2a2a2 — 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/CONCERNS are advisory.

First-Principles-Verdict: CONCERNS

A new 60 s poll and a new dashboardRefetch.ts module add a second refresh path for the drawer, while a slot_projection frame already fires each time the work ledger changes.

Not justified as shipped

  • Item 5 — duplicate of the slot_projection handler in website/src/hooks/useWebSocket.ts:683: the description says "a ledger entry sends no frame that names this query". But the eager folder already publishes a FoldAdvanced event for each work-fold change (crew_log/bus.py), and it reaches the client as a slot_projection frame. That frame names the slot, not the member-panel key.
    Subtraction: delete panelFallbackRefetch, CONDUCTOR_PANEL_TEMPLATE and both refetchInterval hunks in CrewWebview.tsx. Instead, invalidate the observed ['member-panel'] queries inside the existing slot_projection case.
    Clears when: the poll is removed, or the PR shows a work-ledger entry that sends no slot_projection frame to the owner's socket.
  • Item 8 — zero consumers: export { DASHBOARD_FALLBACK_REFETCH_MS } in CrewDynamicDashboard.tsx. Grepping DASHBOARD_FALLBACK_REFETCH_MS finds one importer of the re-export, test/CrewDynamicDashboard.test.tsx, and no app code.
    Subtraction: delete the re-export and import the constant in that test from ./dashboardRefetch.
    Clears when: the re-export is deleted.
  • Item 7 — undeclared: the description says "Published data is read only for next and repo". In the diff, _repo also takes the repo from a ledger item's acceptance.repo.

What this change ships

Inventory (8 items) — 5 justified

Intent: keep the conductor's drawer board current without the conductor having to republish it. FIX (#18956).

  1. Tasks, needs-you and the done count come from the work ledger on every read — justified
  2. A conductor crew with a ledger but no publish now gets a board instead of the empty state — justified
  3. An old next shows muted as "Last plan (time)" — justified
  4. A queued task lights no step — justified
  5. The conductor's drawer re-reads every 60 s while in the foreground — duplicate of the slot_projection frame
  6. The goal-conductor skill no longer asks for a republish each cycle — justified
  7. The board's repo falls back to a pr_checks acceptance's repo — undeclared
  8. The refetch constant moved to a leaf module and is re-exported from the old one — zero consumers

[FIRST-PRINCIPLES-REVIEWED] 45c576d

Description read

[DESCRIPTION-READ] 5984b3db006ac73d001a6c63e94873a7aa308d2b8bd264a992afa7e734fe2f0e

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 Oct 10, 2026 •

Copy link
Copy Markdown
Contributor

UX Review (Opus 5.5) — 🟡 CONCERNS

UX-level review of 45c576d3f804ba7f0cc83f4b63be26cefda2a2a2 — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS/CONCERNS are advisory.

UX-Verdict: CONCERNS

The muted "Last plan" line still tells the reader to wait on a task the board shows as done, so the board seems to contradict itself.

Watch

  • The old plan reads as a current instruction. The template now prints "Last plan (" + from + "): " followed by the agent's old text, and only mutes the colour. In shot-02 the line says "Dispatch T and wait for its PR" right above a T card marked done. The blind reader said it "looks out of date" and "I can't tell whether that line is old or wrong". They also didn't link it to the earlier Next: line: "the wording changed and I don't know why". This shows on every stale cycle, and stale cycles are the case this PR targets. Fix: say outright that the line is out of date (for example "Plan from 17:57, before the latest updates:"), or hide the line once the ledger has items newer than it.
  • The two times can match. In shot-02, "Updated 2026-10-12 17:57 UTC" and "Last plan (2026-10-12 17:57 UTC)" show the same minute, so "Last plan" gives the reader nothing to compare. If no timestamp exists, the fallback "an earlier update" produces "Last plan (an earlier update):". Fix: when both fall in the same minute, show seconds or a relative time ("12 min earlier").
  • The longer timestamp runs into the watermark. Before, the header showed only a time ("15:02 UTC"). stamp() now adds the full date, and in shot-02 the date runs into the faded "Task board" watermark. The reader noticed: "The overlap makes the date harder to read". This shows on every ledger-built board. Fix: show just the time when the date is today, or clip the watermark behind the text.

Suggestions

  • In kirocrew-conductor.html, replace "waiting" with "queued" or "not started". The reader couldn't tell it apart from "needs you": "both sound like something is waiting".

[UX-REVIEWED] 45c576d

Description read

[DESCRIPTION-READ] 60fd5c7283c0f1c058009d75459bb062cfbe0114c3e75bb5069c20f01f6805b7

That sha256 names everything this verdict read: the PR description, plus the 5 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 added readiness: action required A blocking check or review needs attention and removed readiness: checking Automated validation is still running labels Oct 10, 2026
@github-actions

github-actions Bot commented Oct 10, 2026 •

Copy link
Copy Markdown
Contributor

Design Review (Opus 5.5) — 🟡 CONCERNS

Design-level review of 45c576d3f804ba7f0cc83f4b63be26cefda2a2a2 — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS/CONCERNS are advisory.

Design-Verdict: CONCERNS

The skill now says "do not republish", yet a slot shared by two crews still serves only what was published, so that conductor's board never updates.

Watch

  • No board on the published-only path. The skill now reads "draws its tasks … from the work ledger on every read … do not republish it each cycle. panel_publish adds only {next, repo?…}". The handler still builds the ledger board only when owns_ledger = bool(card_slot) and _slug_is_claimed_by_any_member(...), and otherwise "the conductor board is served exactly as published". So a conductor whose slot is shared, or whose slug no configured crew owns, publishes only next/repo and gets a drawer with no tasks, no needs-you and no done count. Before this PR that case was only stale; now it is permanently empty. The skill text and the server's fallback path don't agree.
    Clears when: the skill keeps asking for a full publish whenever the server will serve it as published (for example, the fallback keeps the last published tasks and the skill only stops asking when the ledger board applies), or a test shows a shared-slot conductor still gets a populated board.

Suggestions

  • The ## Backwards compatibility section has both a Compatible: line and a Removes nothing: line. Keep one. The real take-away is that published tasks/needs_you/done are now ignored on read, so say that plainly.

[DESIGN-REVIEWED] 45c576d

Description read

[DESCRIPTION-READ] 494f76f5dc3b015c331639aedd30a26b454f789658b11821f076b1c442946220

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 Oct 10, 2026 •

Copy link
Copy Markdown
Contributor

Opus 5.5 Review — ✅ no blocking findings

Reviewed 45c576d3f804ba7f0cc83f4b63be26cefda2a2a2 — this comment is updated in place on each push.

Review details

No findings that block the merge.

FINDING — docs/system-specs/modules/learn-cron-dashboard.md:2552 — handlers/agent_panel.py changes how the panel read works for kirocrew-conductor, but its owning spec still says only "adds a derived board, which the composer renders in preference to the published data". Readers of the spec are not told that the conductor board now replaces data from the work fold. They are also not told that the read builds a record when nothing was published, that this applies only when owns_ledger holds, or about next_from. → Fix: add a conductor-board bullet to the "Crew member panels" section in this PR, as AGENTS.md requires.

[OPUS-REVIEWED] 45c576d

Verdict parsed from the review's SHA-scoped output markers for commit 45c576d3f804ba7f0cc83f4b63be26cefda2a2a2.

False positive or not applicable? A repository writer can comment:
/ai-review override fable 45c576d3f804ba7f0cc83f4b63be26cefda2a2a2: <one-sentence reason>

@github-actions

github-actions Bot commented Oct 10, 2026 •

Copy link
Copy Markdown
Contributor

GPT 6.1 Review — ✅ no blocking findings

GPT 6.1 completed its review of 45c576d3f804ba7f0cc83f4b63be26cefda2a2a2 and found no blocking issues.

This comment is updated in place on each push.

Review details

No findings.
[GPT-REVIEWED] 45c576d

False positive or not applicable? A repository writer can comment:
/ai-review override gpt 45c576d3f804ba7f0cc83f4b63be26cefda2a2a2: <one-sentence reason>

@github-actions github-actions Bot added the merge conflict Branch has merge conflicts with its base — author must resolve before merge label Oct 11, 2026
@iamwhatever
iamwhatever force-pushed the fix/conductor-panel-fold-board branch from 7740491 to 29b9cd5 Compare October 11, 2026 04:26
@github-actions github-actions Bot added readiness: checking Automated validation is still running readiness: action required A blocking check or review needs attention merge conflict Branch has merge conflicts with its base — author must resolve before merge and removed readiness: action required A blocking check or review needs attention readiness: checking Automated validation is still running merge conflict Branch has merge conflicts with its base — author must resolve before merge labels Oct 11, 2026
@iamwhatever
iamwhatever force-pushed the fix/conductor-panel-fold-board branch from 29b9cd5 to 259293c Compare October 11, 2026 08:13
@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 merge conflict Branch has merge conflicts with its base — author must resolve before merge labels Oct 11, 2026
@iamwhatever
iamwhatever force-pushed the fix/conductor-panel-fold-board branch from 259293c to 5facc97 Compare October 11, 2026 08:40
@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 Oct 11, 2026
@iamwhatever
iamwhatever force-pushed the fix/conductor-panel-fold-board branch from 5facc97 to 84d6b6d Compare October 11, 2026 09:28
@github-actions github-actions Bot removed the readiness: action required A blocking check or review needs attention label Oct 11, 2026
@iamwhatever

Copy link
Copy Markdown
Contributor Author

Dispositions for the CONCERNS on 893742a (author reply):

Design: wrong crews get the conductor board — fixed in 05e4d1d. Verified in code: _panel_record gated the no-publish path on template_for_crew(member) != pipeline_board_contract.BOARD_TEMPLATE_ID, so any crew with a work ledger got a synthesized board. The gate is now positive (== conductor_board_contract.BOARD_TEMPLATE_ID), and test_a_crew_that_is_not_the_conductor_gets_no_board_from_its_ledger pins that another crew with a ledger and no publish reads None (fails on the old gate, passes on the new one).

First Principles: duplicate 60 s constant — accepted, not changed here. PANEL_FALLBACK_REFETCH_MS and DASHBOARD_FALLBACK_REFETCH_MS hold the same value today. Folding them into one shared floor is a small follow-up; it does not change behavior, so this PR keeps its scope.

UX: stale "Next (from )" label — accepted as a wording follow-up. The label already carries the time it was written, which is the honest part; making it say "may be out of date" is copy polish that does not change what the board shows. Left for a follow-up so this PR does not take another full review round.

@github-actions github-actions Bot added the readiness: checking Automated validation is still running label Oct 12, 2026
@iamwhatever
iamwhatever force-pushed the fix/conductor-panel-fold-board branch from be4cf20 to b17ac58 Compare October 12, 2026 10:08
@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 Oct 12, 2026
@iamwhatever

Copy link
Copy Markdown
Contributor Author

Self-review (crew-pr-reviewer, read-only): CHANGES NEEDED

This PR rebuilds the goal conductor's drawer board from the work ledger on every read, so a skipped panel_publish no longer leaves it stale. That fixes the root cause at the right seam. The ledger board is served only when the slot names this crew alone and a configured crew owns the slug, and its text goes through the publish redaction. I tried to read another crew's ledger with a caller-chosen member and could not. The derived data still contradicts itself, and the UX CONCERNS on head b17ac58 has no answer: the author's dispositions are for 893742a. _step (src/kiro_crew/conductor_board_contract.py:141-146) returns "Code" for any item with no PR, including one _state calls "waiting". Before, the conductor wrote the step by hand; now every queued task looks like coding has started. Suggested fix: return an empty step for waiting items (check that the template lights no segment for it) and add a case to test_conductor_board_from_fold.py. The 60 s poll is also wider than the problem. It runs for every crewmate's drawer (website/src/pages/members/CrewWebview.tsx:420 and :1081), though only the conductor template has no frame to refresh it. Its constant also copies DASHBOARD_FALLBACK_REFETCH_MS. Either limit refetchInterval to the conductor template and reuse the existing constant, or post a disposition on this head saying why not. Nit: the from kiro_crew import (pipeline_board_contract,) split in dashboard/handlers/agent_panel.py is noise. — crew-pr-reviewer

@iamwhatever
iamwhatever marked this pull request as draft October 12, 2026 15:48
@iamwhatever

Copy link
Copy Markdown
Contributor Author

Converted to draft by crew-pr-reviewer: the self-review above is CHANGES NEEDED. Mark ready again (gh pr ready 18957) once the points are addressed.

@github-actions github-actions Bot added readiness: checking Automated validation is still running merge conflict Branch has merge conflicts with its base — author must resolve before merge and removed readiness: passed Eligible automated validation passed for the current revision labels Oct 12, 2026
@iamwhatever
iamwhatever force-pushed the fix/conductor-panel-fold-board branch from b17ac58 to 5965892 Compare October 12, 2026 17:12
@github-actions github-actions Bot removed the merge conflict Branch has merge conflicts with its base — author must resolve before merge label Oct 12, 2026
@iamwhatever
iamwhatever marked this pull request as ready for review October 12, 2026 17:28
@iamwhatever

Copy link
Copy Markdown
Contributor Author

Addressed the crew-pr-reviewer self-review and the UX CONCERNS on b17ac58, on head 5965892 (rebased onto main 0cda152, still two commits).

Self-review

  1. _step returned "Code" for a waiting item: fixed. _step now returns an empty step when _state is "waiting"; only an item a worker has started sits on Code. The template's stepIndex("") is -1 and the card is not finished, so no segment is lit and the bar's label reads "Step: not said". New case test_a_waiting_item_has_reached_no_step in test/test_conductor_board_from_fold.py pins waiting -> "", started -> "Code", with a PR -> "PR".
  2. 60 s poll on every crewmate's drawer: fixed. Both member-panel queries in CrewWebview.tsx now use refetchInterval: (query) => panelFallbackRefetch(slug, query.state.data?.panel?.template), which polls only when the record's template is kirocrew-conductor, or the slug is (the server picks a crew's default template by slug, so the conductor's first ledger item still appears in a drawer opened before any board existed). Every other template is back to push-only refresh via panel_published. The copied constant is gone: DASHBOARD_FALLBACK_REFETCH_MS moved to a leaf module website/src/pages/members/dashboardRefetch.ts and CrewDynamicDashboard.tsx re-exports it, so the eager drawer shares it without pulling the lazy Dashboard tab into its chunk (and MembersPage.test.tsx's default-only mock of that module stays valid). Covered by dashboardRefetch.test.ts.
  3. Nit, the split pipeline_board_contract import: folded into the grouped from kiro_crew import (...) block, which is also what ruff's isort order wants.

UX CONCERNS (b17ac58)

  • Waiting tasks look like coding started: same fix as point 1.
  • Stale "Next" reads as the current plan: fixed. When next_from is set the line renders muted (.next.stale) and is labelled "Last plan (2026-10-10 15:02 UTC):" instead of "Next (from …):".
  • Suggestion, same-day HH:MM UTC stamps: not taken in this PR. The full date is what makes a "Last plan" stamp from an earlier day unambiguous, and this PR's scope is the board's source, not its time format.

Verified locally: touched pytest (test_conductor_board_from_fold.py, test_agent_panel_routes.py and every test naming the conductor template or panel templates) 1833 passed; vitest on the members/drawer/dashboard suites 381 passed; tsc --noEmit and eslint clean on the touched files.

@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 Oct 12, 2026
@iamwhatever
iamwhatever force-pushed the fix/conductor-panel-fold-board branch from 5965892 to 54d3d3e Compare October 12, 2026 18:20
@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 Oct 12, 2026
@iamwhatever

Copy link
Copy Markdown
Contributor Author

Disposition for UX Review CONCERNS on 54d3d3e (advisory, no push):

  • Same-minute "Last plan" and "Updated" stamps: the match in shot-02 is an artefact of the capture, where the publish and the ledger entries were written about one second apart. In use, next_from is set only when the ledger moved after the publish, which is a patrol cycle or a worker report later, normally minutes or hours. The muted style and the "Last plan" label carry the "out of date" signal even when the minutes match. Not changed in this PR.
  • Full date in "Updated": kept. The same stamp formatter labels a plan from an earlier day, and a time alone would make that ambiguous. Same-day short stamps are a template-formatting change outside this PR's scope (the board's source).
  • "Earlier plan" wording: noted; "Last plan" sits next to the muted style that marks it as old. Not changed, to avoid another full rerun for a wording suggestion.

Remaining CI red (Backend Tests shard 2, Coverage Gate) is test_wait_discipline_ratchet.py flagging test/test_surface_deny_notice_sites.py and test/test_webhook_permission_requests.py, neither touched here; it reproduces on origin/main 63c155a.

@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 Oct 12, 2026
@iamwhatever
iamwhatever force-pushed the fix/conductor-panel-fold-board branch from 54d3d3e to 4272586 Compare October 12, 2026 18:48
@github-actions github-actions Bot removed the readiness: action required A blocking check or review needs attention label Oct 12, 2026
@iamwhatever

Copy link
Copy Markdown
Contributor Author

Disposition for UX Review CONCERNS on f1ab384 (advisory, no push): these are the same two points answered on 54d3d3e in #18957 (comment), and that answer stands. The matching minutes come from the capture fixture writing the publish and the ledger entries a second apart; in use next_from appears only after a later patrol or worker report. The full-date "Updated" stamp is kept so a plan from an earlier day is unambiguous; a same-day short format is a template-formatting change outside this PR's scope.

Joe Guo added 2 commits October 12, 2026 23:51
The conductor's right-side panel (Dashboard preview off) showed only what the
agent last sent through panel_publish, so a cycle that skipped the publish left
new work items off the board for hours while the work ledger already held them.

The drawer read now rebuilds the kirocrew-conductor board from the same work
fold goal-board reads: tasks, state, PR, the done count and the needs-you band
(an open item whose worker reported question or blocked). A crewmate with a
work ledger that never published gets that board too, on a slot that names it
alone. Published data adds only `next` and `repo`; a `next` older than the
ledger's newest entry carries `next_from` and renders "Next (from <time>)".
The drawer refetches on member_projection and a 60 s foreground poll, the
same liveness the Dynamic Dashboard tab keeps. The goal-conductor skill no
longer asks for a republish each cycle.
The no-publish path built a conductor board for any crew that is not the pipeline conductor, and every crew carries a work ledger. The gate now names the conductor template positively, a test pins that another crew with a ledger and no publish still gets no board, and the feature map names the conductor crew instead of any crewmate.

A queued item that no worker has started reports no step, so the board lights no segment for it; only a started item sits on Code. The drawer's 60 s re-read runs only for the conductor's board (by template, or by slug before any board exists), sharing the Dashboard tab's DASHBOARD_FALLBACK_REFETCH_MS from a leaf module so the lazy tab stays out of the drawer's chunk.

When the published Next line is older than the ledger, the board shows it muted as "Last plan (time):" so it no longer reads as the current instruction.
@iamwhatever

Copy link
Copy Markdown
Contributor Author

Closing as superseded. #19659 (merged to kirodotdev/KiroCrew main on 2026-10-13, commit 1f6f5af) removed the Dynamic Dashboard preview flag and deleted website/src/pages/members/CrewWebview.tsx, the side-panel board this addresses. The Dashboard tab now always renders the goal-board template, which reads the work ledger directly, so a patrol cycle that skips panel_publish can no longer leave that board stale. If a stale board still reproduces on a current Kiro Crew build, please reopen with the version and steps.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

merge conflict Branch has merge conflicts with its base — author must resolve before merge

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Conductor's side-panel task board goes stale when a patrol cycle skips panel_publish

1 participant