agentleFS
Microsoft AzureKnown publisherAzure/osdu-spi-stack/.github/skills/resolve-review-threads/SKILL.md

resolve-review-threads

Skill0 starsChanged 42 days agoApache-2.0View on GitHub

Address PR review feedback by verifying findings, fixing confirmed problems, replying with evidence, and resolving completed threads.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 3 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Verify each finding before editing.

    line 8
  2. Distinctive:

    Inventory: Read unresolved threads and their full discussions using the paginated commands below.

    line 26
  3. Very specific:

    Assess: Read the current code and apply the code-review skill, including governing ADRs.

    line 28
  4. Distinctive:

    Check related occurrences: Look for the same failure in related code and documentation.

    line 32
  5. Very specific:

    Reply: Use gh-voice: state the verdict and supporting evidence.

    line 37

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Resolve Review Threads

Verify each finding before editing. A reviewer's assertion alone is insufficient evidence. Respect the requested scope: assessment-only requests do not authorize edits or GitHub changes; commit and push only when authorized.

Verdicts

VerdictEvidenceAction
ConfirmedObserved behavior or code establishes the claim.Fix the root cause and verify the result.
DisprovedEvidence shows the claim does not apply.Reply with evidence; do not change code for that finding.
InconclusiveEvidence does not establish or refute the claim.State what remains unknown and leave open.

Judge suggestions by evidence and impact, not reviewer type or round. Preference-only suggestions can be declined with a brief reason; they are not disproved defects.

Loop

  1. Inventory. Read unresolved threads and their full discussions using the paginated commands below.
  2. Assess. Read the current code and apply the code-review skill, including governing ADRs. Reproduce nontrivial failures where possible; otherwise state the code evidence and what remains unverified. Failure to reproduce alone does not disprove a finding.
  3. Check related occurrences. Look for the same failure in related code and documentation. Keep fixes within the confirmed issue's scope.
  4. Fix and verify. Address the root cause, not merely the suggested patch. Run relevant checks and uv run pre-commit run --all-files before pushing. Follow repository commit conventions, with one commit per concern.
  5. Reply. Use gh-voice: state the verdict and supporting evidence. Link the pushed fix when applicable; do not describe an unpushed edit as available to the reviewer.
  6. Resolve. Resolve after a verified fix is pushed, a finding is disproved with evidence, or a preference-only suggestion is declined with a reason. Leave inconclusive findings, incomplete fixes, and active discussion open.
  7. Report state. Refresh the paginated inventory and check CI on the latest pushed commit. Report remaining threads, pending CI, or failing checks rather than claiming completion. Leave approval and merge to a human; never enable auto-merge.

Commands

Replace {owner}, {repo}, {pr}, and comment/thread IDs with the target PR's values. Read full discussions from the REST inventory; the GraphQL query fetches only each thread's root comment ID needed to reply.

# Review comments and replies across all pages
gh api --paginate repos/{owner}/{repo}/pulls/{pr}/comments \
  --jq '.[] | {id, in_reply_to_id, path, line, body, user: .user.login}'

# Unresolved threads across all pages
gh api graphql --paginate -f query='
query($endCursor: String) {
  repository(owner: "{owner}", name: "{repo}") {
    pullRequest(number: {pr}) {
      reviewThreads(first: 100, after: $endCursor) {
        pageInfo { hasNextPage endCursor }
        nodes {
          id
          isResolved
          comments(first: 1) { nodes { databaseId } }
        }
      }
    }
  }
}' --jq '.data.repository.pullRequest.reviewThreads.nodes[] | select(.isResolved == false)'

# Reply to a comment
gh api --method POST repos/{owner}/{repo}/pulls/{pr}/comments/{comment_id}/replies \
  -f body='...'

# Resolve a thread
gh api graphql -f query='
mutation { resolveReviewThread(input: {threadId: "{thread_node_id}"}) { thread { isResolved } } }'

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.