Repository navigation
feat: Stamp pr-review and weekly-report with the pinned-SHA footer - #125
Conversation
These programs now follow the Phase 6 attribution rules: a create-only _meta file, a blob/main fallback, and one skill_attribution line on the review body or report. Signed-off-by: Jim Cadden <jcadden@ibm.com> Assisted-by: Cursor <cursoragent@cursor.com>
clawgenti
left a comment
There was a problem hiding this comment.
Adds the pinned-SHA attribution footer to the pr-review flow (new footer command plus startup _meta pinning in the fixer/scanner) and to weekly-report (footer appended to the posted report); the new hermetic test covers create-only _meta, no-refresh of a valid pin, and the blob/main fallback, and all three claimed tests pass locally with CI green and DCO signed.
Findings:
- suggestion —
standing-orders/pr-review.mdstill documents the fixer as "Two modes" (begin/finalize); document the newfootercommand there so the review workflow actually invokes it (see inline comment).
Reviewed by clawgenti using the github-pr-review skill
The standing orders still described only begin and finalize, so the new footer command never reached a review body. Signed-off-by: Jim Cadden <jcadden@ibm.com> Assisted-by: Cursor <cursoragent@cursor.com>
|
Addressed the standing-orders suggestion in 1bb5706.
Assisted-By: Claude Code |
clawgenti
left a comment
There was a problem hiding this comment.
The follow-up commit addresses the prior docs gap — standing-orders/pr-review.md now documents the fixer's footer command and where the pinned-SHA line goes — with CI green on both legs, DCO signed, and the new hermetic attribution test wired into both the shellcheck and run lists.
All checks pass. Ready for human review.
Reviewed by clawgenti using the github-pr-review skill
The 6a tag only makes sense next to the design spec. Standing orders stay untouched for the run-path migration. Signed-off-by: Jim Cadden <jcadden@ibm.com> Assisted-by: Cursor <cursoragent@cursor.com>
|
Took the human review in 5982a35.
Assisted-By: Claude Code |
5982a35 to
cbda6ab
Compare
clawgenti
left a comment
There was a problem hiding this comment.
Follow-up on 5982a35: this docs-only commit removes the 6a phase tags from the provenance comments, states the footer rules directly in the new test's header, and reverts the standing-orders edit in favor of #118 as requested. The reviewed behavior is unchanged, DCO is signed on all three commits, and CI is green on both legs.
All checks pass. Ready for human review.
Reviewed by clawgenti using the github-pr-review skill
clawgenti
left a comment
There was a problem hiding this comment.
Follow-up on 5982a35: docs-only push restating the footer rules without phase labels in the script comments and the new test header; reviewed behavior is unchanged, CI is green on both legs, and DCO is signed on all three commits.
All checks pass. Ready for human review.
Reviewed by clawgenti using the github-pr-review skill
Summary
pr-review-fixer.sh footerprints thegithub-pr-reviewattribution line for the review body and its Slack post. Inline comments stay without it. The scanner and fixer pin_metaon startup: write it only after the SHA resolves, never overwrite a valid file, and fall back toblob/mainwhen resolution fails.weekly-report.shappends thegithub-weekly-reportline to the report that gets posted.Refs rossoctl/repoman#10
Test plan
bash tests/test-attribution-consumers.shbash tests/test-weekly-report.shbash tests/test-skill-meta.shAssisted-By: Claude Code