Skip to content

docs(rfc): conductor skips the in-chat task board on a board tab - #18620

Merged
iamwhatever merged 1 commit into
mainfrom
docs/rfc-conductor-chat-board-dedupe
Oct 10, 2026
Merged

iamwhatever merged 1 commit into
mainfrom
docs/rfc-conductor-chat-board-dedupe

Conversation

@iamwhatever

@iamwhatever iamwhatever commented Oct 9, 2026 •

Copy link
Copy Markdown
Contributor

Problem / Motivation

Goal: Record the decision that a conductor whose Dashboard tab shows the task board stops drawing a second full copy of it in the chat, so its implementation can land.

The First Principles lane blocked #18608 as a product-shape change with no RFC on the base branch: it changes what kirocrew-conductor draws in chat by default on one surface. Crew Mode users reported the in-chat "Task board" at every milestone as a duplicate of the Dashboard tab's goal-board.

Why it matters

Without the decision on main the implementation cannot pass First Principles, and Crew Mode chats keep repeating a full table the reader already has one tab away.

Not a goal

What changed (motivation → approach → change)

  • docs/request-for-change/rfc-conductor-chat-board-on-tab.md: the RFC, status in-progress, implementation #18608. It states the problem, the two-branch rule (keyed on the [DASHBOARD] block naming goal-board / work-kanban), why "a block exists" is the wrong key (project-report fallback), the task-dashboard artifact disposition, risks, and how it squares with Bolin Chen's recorded decision.
  • docs/request-for-change/README.md: one index row.

@bolichen97 — this records a change to the Crew Mode milestone rule you added in #16088. Your review is requested.

Backwards compatibility

Compatible: documentation only.

Removes nothing: adds one RFC file and one index row.

Tests

N/A — docs only; scripts/docs-lint.sh passes locally.

Manual verification

N/A — docs only.

Related Issues

Implementation: #18608. Follows #16088.

Checklist

  • At most two commits (one is the norm), with a Conventional Commits title
  • 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 9, 2026 21:31
@iamwhatever
iamwhatever requested a review from bolichen97 October 9, 2026 21:31
@iamwhatever

Copy link
Copy Markdown
Contributor Author

@bolichen97 could you take a look when you have a moment? This RFC narrows the Crew Mode milestone rule you added in #16088, so it needs your on-record OK before it lands.

The reason comes from Crew Mode users: the conductor's Dashboard tab (goal-board / work-kanban) already draws the full task board, live from the work ledger, so the full in-chat "Task board" at every milestone shows every task twice.

What stays the same:

  • The intro widget on the first reply.
  • "Needs you" still comes first in the chat at every milestone, plus one line on what changed and what is next.
  • Every surface without a board tab (other templates, channels, cron, CLI) keeps the full in-chat board.

Section 5 explains why I read this as keeping your 2026-10-02 decision (the tab is the live board). The review lanes read it as a narrowing that needs a superseding entry. If you agree with the change, a short "OK" here is enough, and I will add that superseding decision entry citing your comment. If you would rather keep the in-chat board everywhere, I will close both PRs. Thank you.

@github-actions github-actions Bot added the readiness: checking Automated validation is still running label Oct 9, 2026
@github-actions

github-actions Bot commented Oct 9, 2026 •

Copy link
Copy Markdown
Contributor

UX Review (Opus 5.5) — ⏭️ skipped

Revision abc9aea274acc3365ba59b84643e6a6fac4a0a5f touches no user-facing surface (no changes under website/ or committed screenshots), so the UX review was skipped. Advisory — does not block merge.

@github-actions

github-actions Bot commented Oct 9, 2026 •

Copy link
Copy Markdown
Contributor

Design Review (Opus 5.5) — 🟡 CONCERNS

Design-level review of abc9aea274acc3365ba59b84643e6a6fac4a0a5f — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS/CONCERNS are advisory.

Design-Verdict: CONCERNS

The RFC decides for itself that it fits Bolin Chen's recorded decision ("reports with a live task board … in every chat"), yet narrows that rule without the maintainer signing off.

Watch

  • Self-certified fit with a recorded decision. Section 5 says "This change keeps it", but the header says it "Narrows one prompt rule" that docs/decisions/2026-10-02-conductor-introduces-itself-in-every-chat.md records. Under AGENTS.md, only a superseding entry written by a maintainer can reverse or narrow a recorded decision; the author's own reading can't. So the implementation in fix(conductor): skip the in-chat task board when the Dashboard tab shows it #18608 would rest on a reinterpretation nobody has approved. Chat-only readers would lose the per-task rows on board tabs, and a later reviewer could fairly call that a decision violation and revert it. The RFC already lists the surface-scoping as safe, and I checked it against the code: turn_block is only injected on desk turns, because desk_withheld covers cron, channel and delegated turns, and a failed read returns "".
    Clears when: @bolichen97 approves on this PR that "the tab is the live board" meets the decision, or a maintainer adds a superseding decision entry that narrows it.

[DESIGN-REVIEWED] abc9aea

Description read

[DESCRIPTION-READ] bf7aaa9dc4ec33c75dc8fb122e6ed1646d9f0073c120f96b0349797a07622367

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 9, 2026 •

Copy link
Copy Markdown
Contributor

First Principles Review (Opus 5.5) — ⏭️ skipped

Revision abc9aea274acc3365ba59b84643e6a6fac4a0a5f ships no reviewable capability (docs, tests or generated files only), so there is nothing to inventory. Advisory — does not block merge.

@github-actions

github-actions Bot commented Oct 9, 2026 •

Copy link
Copy Markdown
Contributor

Opus 5.5 Review — ✅ no blocking findings

Reviewed abc9aea274acc3365ba59b84643e6a6fac4a0a5f — this comment is updated in place on each push.

Review details

No findings.

[OPUS-REVIEWED] abc9aea

Verdict parsed from the review's SHA-scoped output markers for commit abc9aea274acc3365ba59b84643e6a6fac4a0a5f.

False positive or not applicable? A repository writer can comment:
/ai-review override fable abc9aea274acc3365ba59b84643e6a6fac4a0a5f: <one-sentence reason>

@github-actions

github-actions Bot commented Oct 9, 2026 •

Copy link
Copy Markdown
Contributor

GPT 6.1 Review — ✅ no blocking findings

GPT 6.1 completed its review of abc9aea274acc3365ba59b84643e6a6fac4a0a5f and found no blocking issues.

This comment is updated in place on each push.

Review details

No findings.
[GPT-REVIEWED] abc9aea

False positive or not applicable? A repository writer can comment:
/ai-review override gpt abc9aea274acc3365ba59b84643e6a6fac4a0a5f: <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 9, 2026
@iamwhatever
iamwhatever force-pushed the docs/rfc-conductor-chat-board-dedupe branch from 6c8d4ca to abc9aea Compare October 10, 2026 06:03
@iamwhatever

Copy link
Copy Markdown
Contributor Author

Note for reviewers, from the user who raised this: the point is to NOT GENERATE the board, which saves output tokens, not to hide one after it is drawn.

The rule is a generation-time rule in the conductor prompt. On a board tab (goal-board / work-kanban) the model never writes the "Task board" widget or the task-dashboard artifact, so there is no client-side suppression, filter or hide anywhere. The motivation is both costs: the reader sees every task twice, and each milestone spends output tokens on a widget plus an artifact update the tab already shows. Section 3 now says this explicitly (head abc9aea). The intro widget and the "Needs you" line stay.

Still waiting on @bolichen97's on-record OK before the implementation (#18608) moves.

@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 10, 2026
@iamwhatever
iamwhatever merged commit 3f9dd33 into main Oct 10, 2026
111 of 115 checks passed
@iamwhatever
iamwhatever deleted the docs/rfc-conductor-chat-board-dedupe branch October 10, 2026 18:36
@github-actions github-actions Bot removed the readiness: action required A blocking check or review needs attention label Oct 10, 2026
@iamwhatever

Copy link
Copy Markdown
Contributor Author

This is a RFC doc, also synced offline about this change to avoid duplicate dashboard.

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.

1 participant