Add unresolved PR review skill - #47
Conversation
Introduce an engineering workflow for evaluating every actionable unresolved review comment, recording explicit decisions with reactions, and posting suggested replies only after exact user approval. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Keep generated Python cache files out of the engineering skill package. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
🦋 Changeset detectedLatest commit: d0a075f The changes in this PR will be included in the next version bump. This PR includes changesets to release 1 package
Not sure what this means? Click here to learn what changesets are. Click here if you're a maintainer who wants to add another changeset to this PR |
|
=== review-pr-comments === Running benchmark: review-pr-comments-eval ✓ [1/3] Review Agent Feedback 🧪 Waza Eval ResultsStatus: ✅ Passed | Score: 1.00 | Duration: 229ms
Task Results
Benchmark: review-pr-comments-eval | Skill: engineering/review-pr-comments | Model: claude-sonnet-4-20250514 === branch-name === Running benchmark: branch-name-eval ✓ [1/3] Feature Branch Naming 🧪 Waza Eval ResultsStatus: ✅ Passed | Score: 1.00 | Duration: 229ms
Task Results
Benchmark: branch-name-eval | Skill: git/branch-name | Model: claude-sonnet-4-20250514 === commit-message === Running benchmark: commit-message-eval ✓ [1/3] Basic Bug Fix Commit 🧪 Waza Eval ResultsStatus: ✅ Passed | Score: 1.00 | Duration: 230ms
Task Results
Benchmark: commit-message-eval | Skill: git/commit-message | Model: claude-sonnet-4-20250514 === pr-title-description === Running benchmark: pr-title-description-eval ✓ [1/3] Basic Feature PR 🧪 Waza Eval ResultsStatus: ✅ Passed | Score: 1.00 | Duration: 232ms
Task Results
Benchmark: pr-title-description-eval | Skill: git/pr-title-description | Model: claude-sonnet-4-20250514 ═══════════════════════════════════════════════ Skill Status Pass Rate Avg Score |
Remove the Python helper and document direct gh GraphQL and REST commands for paginated collection, decision reactions, and approval-gated replies. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
|
=== review-pr-comments === Running benchmark: review-pr-comments-eval ✓ [1/3] Review Agent Feedback 🧪 Waza Eval ResultsStatus: ✅ Passed | Score: 1.00 | Duration: 206ms
Task Results
Benchmark: review-pr-comments-eval | Skill: engineering/review-pr-comments | Model: claude-sonnet-4-20250514 === branch-name === Running benchmark: branch-name-eval ✓ [1/3] Feature Branch Naming 🧪 Waza Eval ResultsStatus: ✅ Passed | Score: 1.00 | Duration: 206ms
Task Results
Benchmark: branch-name-eval | Skill: git/branch-name | Model: claude-sonnet-4-20250514 === commit-message === Running benchmark: commit-message-eval ✓ [1/3] Basic Bug Fix Commit 🧪 Waza Eval ResultsStatus: ✅ Passed | Score: 1.00 | Duration: 207ms
Task Results
Benchmark: commit-message-eval | Skill: git/commit-message | Model: claude-sonnet-4-20250514 === pr-title-description === Running benchmark: pr-title-description-eval ✓ [1/3] Basic Feature PR 🧪 Waza Eval ResultsStatus: ✅ Passed | Score: 1.00 | Duration: 206ms
Task Results
Benchmark: pr-title-description-eval | Skill: git/pr-title-description | Model: claude-sonnet-4-20250514 ═══════════════════════════════════════════════ Skill Status Pass Rate Avg Score |
Why
Unresolved pull-request feedback needs a consistent workflow that includes agent-authored comments, records explicit decisions, and prevents accidental replies.
What
engineering-skillsmarketplace plugin andreview-pr-commentsskillgh api graphqlcallsgh apiHow to verify
gh api graphqlcollection query against a pull requestwaza checkand thereview-pr-commentsevaluation suiteRisks / follow-ups
ghsession with access to the target repository