feat(skills): enhance pr-status-all and pr-status with hyperlinked review verdicts and review summary table - #2097
Conversation
…view verdicts and review summary table Closes #2096
|
@claude review |
|
👀 Picked up by workflow run #32696755295. Setup runs first; Claude itself responds after that. |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
- Restore operational rationale for external review and human review checks - Restore graceful degradation to series section - Factor human CHANGES_REQUESTED, threads, behind-main, and draft into Next Step - Use generic placeholders for example table rows - Mark draft PRs in assemble step and output
ARD Response — Round 1
@claude review |
|
👀 Picked up by workflow run #32697466397. Setup runs first; Claude itself responds after that. |
![claude[bot]](https://avatars.githubusercontent.com/in/1236702?s=60&v=4){kind=small}
| - **Next Step** --- computed deterministically using the full state matrix: | ||
| - If `isDraft`: `Draft (Work in progress)`. | ||
| - If human `CHANGES_REQUESTED` is pending: `Blocked on human changes (<login>)` (overrides everything below). | ||
| - If branch is behind main: `Resolve conflicts / Sync main (<N> commits behind)`. | ||
| - If CI is failing: `Fix CI (<failing-check>)`. | ||
| - If unaddressed review threads remain: `Resolve inline threads (<N> open)`. | ||
| - If AI review has open findings: `Drive to clean (ARDI)`. | ||
| - If AI review is running: `In-flight AI review`. | ||
| - If fully clean (no human blocks, AI/external review clean, CI green, 0 open threads, up to date with main): | ||
| - If `Author` is `d-morrison` (self-authored): `Ready for self-merge`. | ||
| - If `Author` is external and human review is requested (`d-morrison`): `Ready for human review`. | ||
| - If `Author` is external and human review is not yet requested: `Request human review`. |
There was a problem hiding this comment.
"Next Step" here depends on two signals this file never gathers.
This decision tree branches on isDraft (line 333) and "branch is behind main" (line 335), but unlike pr-status-all/SKILL.md (which fetches isDraft in its LIST_PRS call and computes behind-main via git fetch origin main <headRefName> -q && git rev-list --count origin/<headRefName>..origin/main in item 6), this single-PR file has no step anywhere that fetches either value — the top gh pr view <N> --json state,title (line 41) only reads state/title, and there is no git fetch/rev-list command in the file at all.
A session following this skill literally has no data to evaluate the isDraft and "behind main" branches of the Next Step matrix. Add a fetch step for both (e.g. extend the top gh pr view call to include isDraft, and add the same git fetch origin main <branch> -q && git rev-list --count check pr-status-all uses), or drop those two branches from this file's Next Step logic if they're intentionally out of scope for a single-PR check.
|
|
||
| When detailed git/thread metrics are needed, include the extended columns: | ||
|
|
||
| | PR | Title | Branch | CI | Review | External | Human | Threads | Behind main | Next Step | |
There was a problem hiding this comment.
Missing header-separator row — this won't render as a Markdown table.
GFM tables need a delimiter row (|---|---|...|) immediately below the header row to render as a table; this header is followed directly by a blank line and then prose. Compare with the Primary Review Summary Table just above, which correctly has |:---|:---|:---:|:---:|:---:|:---| under its header (line 196). As written, this line will render as plain text with pipe characters rather than a table, in any client that renders this skill's Markdown.
| | PR | Title | Branch | CI | Review | External | Human | Threads | Behind main | Next Step | | |
| | PR | Title | Branch | CI | Review | External | Human | Threads | Behind main | Next Step | | |
| |:---|:---|:---|:---:|:---:|:---:|:---:|:---:|:---:|:---| |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
- Fetch isDraft, headRefName, and behind-main check in pr-status - Add markdown delimiter row to Extended Technical Dashboard table in pr-status-all - Add companion pr-status-all.rationale.md documenting operational rationale and synthetic fixtures
ARD Response — Round 2
@claude review |
|
👀 Picked up by workflow run #32698042477. Setup runs first; Claude itself responds after that. |
…ive sentences in pr-status-all
|
💰 Cost: $1.0483728 |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
ARD Response — Round 6
@claude review |
|
👀 Picked up by workflow run #32700399228. Setup runs first; Claude itself responds after that. |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
…athering and output specs
ARD Response — Round 7
@claude review |
|
👀 Picked up by workflow run #32700876347. Setup runs first; Claude itself responds after that. |
2026-08-24 07:53 UTC |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
|
Claude finished review — View run Code Review — PR #2097 (confirming review, head commit
|
1 participant
Summary
Enhance \skills/pr-status-all/SKILL.md\ and \skills/pr-status/SKILL.md\ to produce a clean Review Summary Table:
Closes #2096