agentleFS
asarelcastellanosasarelcastellanos/claude-workbench/skills/address-review-findings/SKILL.md

address-review-findings

Skill1 starsChanged 4 days agoMITView on GitHub

Use when you respond to code review findings on your own change. Every finding gets an answer, fixes cover the whole class of mistake, and the checks run again.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 3 are distinctive, and 1 is 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.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this when a reviewer reports findings on a change you wrote.

    line 8
  2. Distinctive:

    Check each finding before you fix it.

    line 19
  3. Distinctive:

    Before you fix a finding, search the change and nearby code for the same mistake.

    line 23
  4. Very specific:

    After you fix, run the checks from the verify-before-pr skill again.

    line 33

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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

Address review findings

Use this when a reviewer reports findings on a change you wrote. The reviewer can be a person or an agent.

1. Answer every finding

Give each finding one outcome. Never skip a finding.

  • Fixed. Say what changed.
  • Rejected as wrong. Show the evidence, such as the line that handles the case or a test that proves it.
  • Intentional. Explain why the behavior is wanted.
  • Doesn't matter. Explain why the impact is small enough to ignore.

Check each finding before you fix it. Reviewers make mistakes. Do not apply a fix you believe is wrong just to close the comment.

2. Fix the whole class

Before you fix a finding, search the change and nearby code for the same mistake. Fix every instance in one pass, so no copy of the bug stays behind.

Fixes go to a model-policy:coder, because the main session cannot edit code files. The main session folds each fix into the existing commit it belongs to, as the "Commit grouping" section of git-messages describes. A fix never gets a new PR.

3. Handle pre-existing findings

A pre-existing finding is a bug the change did not introduce. Fix it only if the fix is small and fits the scope of the change. Otherwise list it as a follow-up for the user. Do not let it grow the PR.

4. Verify again

After you fix, run the checks from the verify-before-pr skill again. That means the tests, type checks, linter, build, each issue's "Done when" checks, and the end-to-end check. Show the output.

The review does not run again, except in the case step 6 of ship-a-change names: a finding that shows the approach is wrong, or a fix that grows past the size limit there or leaves the reviewed diff. Then the plan goes back to the user, and only the fix's diff is reviewed.

5. Report

End with a table of every finding and its outcome. Then give the check results and any follow-ups. Mistakes worth remembering go into LESSONS.md at step 7 of ship-a-change, not here.

FindingTagOutcomeNote

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.