agentleFS
amkiskoamkisko/prayers/.agents/skills/change-review/SKILL.md

change-review

Skill2 starsChanged 10 days agoMITView on GitHub

Review a git diff with file coverage, path-matched criteria, and falsify-before-report. Use when asked to review a pull request, a commit, or a working-tree diff. Do not use for a whole-tree engineering audit, session retrospective, or live service intake.

10 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 4 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 9 edited versions we read keep these.

  • 4

    Distinctive

    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:

    Read the file. Cite line ranges that exist in the current tree.

    line 24
  2. Distinctive:

    Use path-matched criteria.

    line 26
  3. Distinctive:

    Do not restate what formatters, linters, and compilers already catch.

    line 32
  4. Very specific:

    Use observed, inference, or artifact-missing as the evidence kind.

    line 36
  5. Very specific:

    Write durable findings under docs/issues when the review is more than a chat pass.

    line 40
  6. Distinctive:

    When a named skill is unavailable, preserve the finding and name the deeper review.

    line 49

“Common” is measured: kept by at least 6 in 10 of the 9 edited versions of this file we could read. “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

Change review

Review the changed files in a diff. This skill does not replace engineering-audit. It owns coverage of the change, not the whole tree.

Quick reference

list changed files -> mark skips -> path-matched criteria -> read the file -> cite existing lines -> falsify against the diff -> report

Coverage

Account for every changed file. Mark skipped files with a reason. Report coverage. Do not stop after the first high finding.

Read the file. Cite line ranges that exist in the current tree. A finding that cannot be placed is inference or artifact-missing.

Use path-matched criteria. Do not dump every language checklist into one prompt.

Falsify

Falsify each finding against the diff alone before reporting. Drop a finding the diff contradicts. Confirm concurrent invocation before a race. Confirm attacker control before a security finding.

Do not restate what formatters, linters, and compilers already catch. Missing tests remain a defect.

Kinds

Use observed, inference, or artifact-missing as the evidence kind. When security mode is also in scope, add the disposition defined by engineering-audit security mode.

Record

Write durable findings under docs/issues when the review is more than a chat pass. Follow docs-conventions. The chat report may be compact; the file stays plain prose.

Routing

  • engineering-audit for whole-tree, pipeline, or product-mode review.
  • claims-audit when the diff states material external facts.
  • publish-checks before a pull request.
  • session-review for transcripts, not diffs.

When a named skill is unavailable, preserve the finding and name the deeper review.

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.