Skip to content

fix(dashboard): the conductor page counts the tasks the work fold forgot - #19902

Merged
chenmingwei23 merged 1 commit into
mainfrom
fix/conductor-page-forgotten
Oct 14, 2026
Merged

chenmingwei23 merged 1 commit into
mainfrom
fix/conductor-page-forgotten

Conversation

@iamwhatever

@iamwhatever iamwhatever commented Oct 13, 2026 •

Copy link
Copy Markdown
Contributor

Problem / Motivation

Goal: The conductor's default dashboard page shows true done and total counts on a long goal.

The builtin kirocrew-conductor page (#19706) counts done and total from the work fold's items only. Since #19068 the work fold forgets closed items past its window: they leave items and are counted in work.forgotten. So once old tasks are forgotten, the page's meter undercounts both numbers. goal-board already reads forgotten; this page does not.

Why it matters

Every conductor crewmate opens this page by default. On a goal with more closed tasks than the fold keeps, the meter says far less is done than really is.

Not a goal

  • No change to the work fold or to how it forgets items.
  • No change to goal-board or any other builtin page.
  • No new list of forgotten tasks: the page still draws cards only for what items holds.

What changed (motivation → approach → change)

Symptom: done and total shrink once the fold forgets closed tasks. Cause: the page never bound work.forgotten. Change: src/kiro_crew/dashboard_templates/builtin/kirocrew-conductor/manifest.json binds {"fold": "work", "path": "forgotten"} the way goal-board does, and the meter in its template.html adds it in the way goal-board (drawLoop) and pipeline_board_contract.py count a board, so one goal reads the same total on every page:

  • done = listed accepted + forgotten.accepted
  • total = listed + forgotten.items (rejected and abandoned tasks stay in, as before this PR)

The total no longer adds omitted: that counter counts dropped log entries, not tasks (pipeline_board_contract.py says the same and leaves it out), so adding it made a total no other page shows. The truncation line under the cards still reads omitted. A missing or odd forgotten (an older gateway) reads as zeros. The manifest version stays 1, as #19068 left goal-board's.

Backwards compatibility

Compatible: the new field is read-only and optional; a gateway without forgotten gives zeros. The one visible change on such a gateway is that a board with omitted > 0 no longer adds those entries to its total, matching goal-board.

Removes nothing: the change only adds a bound field and adds counts to the meter.

Tests

  • website/src/test/conductorBoardPage.test.ts: a forgotten-heavy fixture (2 listed, 11 forgotten) reads "9 of 13", the same total goal-board gives, with matching progressbar ARIA values; a listed rejected task stays in the total and is not counted as working; omitted entries stay out of the total; missing, null, string, array and negative forgotten values read as zeros.
  • test/test_dashboard_template_kirocrew-conductor.py: forgotten is bound exactly as goal-board binds it, and the meter reads forgotten.items and forgotten.accepted.
  • test/test_dashboard_templates_builtin.py (unchanged) stays green: the new path resolves in the captured work fold with the declared type.

Determinism: the new vitest cases are pure DOM over a fixed field bag; the file ran green 3 times in a row and in shuffled order.

Manual verification

Rendered the page before and after in headless Chromium on the same forgotten-heavy field bag (sample data; the host's own field fill is not simulated, so the goal reads "-").

Screenshots / video

Rendered in headless Chromium on the forgotten-heavy fixture (2 listed, 11 forgotten), dark theme.

Before (main): "1 of 2 accepted by the conductor", bar at 50%.

Before: 1 of 2 accepted, bar at 50%

After: "9 of 13 accepted by the conductor · 1 working", bar at 69%.

After: 9 of 13 accepted, bar at 69%

The cards and the rest of the page are unchanged.

Related Issues

Closes #19886

Follow-up to #19706 and #19068.

Pattern harvest

Rule candidate: review-prompt
Pattern: a builtin page that draws a whole-board count from work.items without adding work.forgotten. Known sibling left open: flow's "accepted … of N" tile (#20023).

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

@iamwhatever
iamwhatever requested review from a team and KiroCrewCaptain as code owners October 13, 2026 19:26
@github-actions github-actions Bot added the readiness: checking Automated validation is still running label Oct 13, 2026
@github-actions

github-actions Bot commented Oct 13, 2026 •

Copy link
Copy Markdown
Contributor

First Principles Review (Opus 5.5) — 🟡 CONCERNS

Premise-level review of b036774e515d93720eb22b056c4c416518f0f15d — 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

The pull request (PR) fixes the conductor meter, but the same undercount is still live in flow's "accepted … of N" tile, so "Not generalizable" doesn't hold.

Not justified as shipped

  • Item 1 — symptom-level (point patch, 1 unfixed sibling): the description says "Not generalizable: one page missed one field a sibling page already binds". 8 builtin manifests bind "fold": "work", "path": "items" and 5 of them don't bind forgotten. Of those 5, flow/template.html:265 builds a whole-board number from items alone: i18n('kpi_of', { n: list.length }). It undercounts exactly like the conductor meter did before this PR. The cause is that every page counts the board its own way. forgotten() now has 3 hand-written copies (goal-board, work-kanban, kirocrew-conductor). ## Not a goal excludes other pages, so the fix is deferred, but the description's claim is wrong.
    Clears when: the description drops "Not generalizable" and names flow's accepted tile as a known sibling left open.

What this change ships

Inventory (4 items) — 3 justified

Intent: the conductor's default page shows true done and total counts on a long goal. FIX.

  1. The meter's done and total now include the closed tasks the fold forgot — symptom-level, flow sibling left unfixed
  2. The meter's total no longer adds omitted (dropped log entries) — justified
  3. A missing or malformed forgotten reads as zero — justified
  4. The manifest description mentions the forgotten block — justified

[FIRST-PRINCIPLES-REVIEWED] b036774

Description read

[DESCRIPTION-READ] 58da5ddadcef383618f1045f0d3287b99963b5a44028f35201850f8021b5d0c5

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 13, 2026 •

Copy link
Copy Markdown
Contributor

UX Review (Opus 5.5) — 🟡 CONCERNS

UX-level review of b036774e515d93720eb22b056c4c416518f0f15d — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS/CONCERNS are advisory.

UX-Verdict: CONCERNS

The meter now says "9 of 13" above only two cards, and nothing on the page explains where the other 11 tasks went.

Watch

  • The count and the card list disagree. The new total is list.length + gone.items, but cards are still drawn only from items, and the truncation line (#cb-more) only appears when omitted > 0. The blind reader, on shot-02, said: "the page says there are 13 tasks and only 2 are listed. I can't see the other 11… I can't tell… whether the count or the list is wrong." This hits every conductor crewmate on any long goal, every time they open their default page. It makes the corrected number look like a bug. Fix: when forgotten().items > 0, show a muted line under the cards, e.g. "11 earlier closed tasks not listed". That is a count, not a list, so it stays inside the PR's "Not a goal".

Suggestions

  • more: 'more tasks not shown' is still driven by omitted. The PR itself says omitted counts dropped log entries, not tasks, so this line now labels entries as "tasks". Reword it to match what it counts, or fold it into the line above.

[UX-REVIEWED] b036774

Description read

[DESCRIPTION-READ] 4140105fd3ed3ffb832731edf809702cc13210cf8d1f822389e49c103a1a6637

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 commented Oct 13, 2026 •

Copy link
Copy Markdown
Contributor

Design Review (Opus 5.5) — ✅ PASS

Design-level review of b036774e515d93720eb22b056c4c416518f0f15d — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS/CONCERNS are advisory.

Design-Verdict: PASS

Nothing to check.

[DESIGN-REVIEWED] b036774

Description read

[DESCRIPTION-READ] 8f10b50e5404c9eb3a5a5493d0e27fdfd00a69d54b442b3a9f35ea7695b7a74b

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 13, 2026 •

Copy link
Copy Markdown
Contributor

GPT 6.1 Review — ✅ no blocking findings

GPT 6.1 completed its review of b036774e515d93720eb22b056c4c416518f0f15d and found no blocking issues.

This comment is updated in place on each push.

Review details

No findings.
[GPT-REVIEWED] b036774

False positive or not applicable? A repository writer can comment:
/ai-review override gpt b036774e515d93720eb22b056c4c416518f0f15d: <one-sentence reason>

@github-actions

github-actions Bot commented Oct 13, 2026 •

Copy link
Copy Markdown
Contributor

Opus 5.5 Review — ✅ no blocking findings

Reviewed b036774e515d93720eb22b056c4c416518f0f15d — this comment is updated in place on each push.

Review details

No findings.

[OPUS-REVIEWED] b036774

Verdict parsed from the review's SHA-scoped output markers for commit b036774e515d93720eb22b056c4c416518f0f15d.

False positive or not applicable? A repository writer can comment:
/ai-review override fable b036774e515d93720eb22b056c4c416518f0f15d: <one-sentence reason>

@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 13, 2026
@iamwhatever
iamwhatever force-pushed the fix/conductor-page-forgotten branch from 96cf2e0 to c958ce0 Compare October 13, 2026 20:00
@github-actions github-actions Bot added readiness: checking Automated validation is still running readiness: action required A blocking check or review needs attention and removed readiness: action required A blocking check or review needs attention readiness: checking Automated validation is still running labels Oct 13, 2026
@iamwhatever
iamwhatever force-pushed the fix/conductor-page-forgotten branch from c958ce0 to ab0228b Compare October 13, 2026 22:42
@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 13, 2026
@iamwhatever
iamwhatever force-pushed the fix/conductor-page-forgotten branch from ab0228b to b036774 Compare October 13, 2026 23:37
@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 14, 2026
@iamwhatever

Copy link
Copy Markdown
Contributor Author

span=0d3060dd15a6
fixed (description, at head b036774)

Confirmed: flow/template.html:261-265 counts its accepted tile from list.length and never binds forgotten, the same undercount.
The description's Pattern harvest no longer says "Not generalizable". It names the pattern as a review-prompt rule candidate and flow as a known sibling left open.
flow is tracked as #20023, outside this PR's one-page scope (Not a goal: no change to any other builtin page).

@iamwhatever

Copy link
Copy Markdown
Contributor Author

span=4c6ae75862e8
rebutted

The meter counting more tasks than cards is the fix, not a regression: the work fold forgets closed tasks by design (#19068) and the board's total must still include them.
The sibling page ships the same shape: goal-board/template.html:335 (n: list.length + gone.items) and :456 show the whole-board count over only the listed cards, with no extra line.
A new "N closed tasks not listed" line would be a new surface plus i18n on one page and not its sibling; the task scope is one mechanism (count the forgotten items).

@iamwhatever

Copy link
Copy Markdown
Contributor Author

span=19deb628ce4f
rebutted

Not changed by this PR: more: 'more tasks not shown' (kirocrew-conductor/template.html:109) and the #cb-more line that reads omitted are both on main as #19706 shipped them; this diff touches neither.
The PR only stops adding omitted into the meter total, which is what makes the meter match goal-board.
Rewording the truncation line is a separate copy change to a line this PR leaves alone.

@github-actions github-actions Bot added readiness: passed Eligible automated validation passed for the current revision and removed readiness: action required A blocking check or review needs attention labels Oct 14, 2026
@chenmingwei23
chenmingwei23 merged commit 1b3a2c1 into main Oct 14, 2026
112 checks passed
@chenmingwei23
chenmingwei23 deleted the fix/conductor-page-forgotten branch October 14, 2026 01:40
@github-actions github-actions Bot removed the readiness: passed Eligible automated validation passed for the current revision label Oct 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.

Conductor dashboard page undercounts done/total once the work fold forgets closed items

2 participants