agentleFS
refpathrefpath/chatgpt-team/.agents/skills/independent-review/SKILL.md

independent-review

Skill0 starsChanged 2 months agoMITView on GitHub

Review a candidate change without modifying it, prioritizing concrete correctness and compatibility findings over stylistic preferences.

14 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

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

  • 0

    Common

    Most of the 10 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read applicable AGENTS.md review rules.

    line 11
  2. Distinctive:

    Check correctness, failure paths, boundary cases, concurrency/resource behavior, compatibility, unintended scope, and test adequacy.

    line 13
  3. Distinctive:

    If no consequential issue is found, state what was reviewed and what remains unverified.

    line 35

“Common” is measured: kept by at least 6 in 10 of the 10 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

Independent review

Procedure

  1. Pin the review to an exact candidate SHA or immutable diff.
  2. Read applicable AGENTS.md review rules.
  3. Inspect the diff and enough surrounding code to understand behavior and contracts.
  4. Check correctness, failure paths, boundary cases, concurrency/resource behavior, compatibility, unintended scope, and test adequacy.
  5. Compare implementation against acceptance criteria rather than the author's narrative.
  6. Validate important concerns with code/tests/commands when read-only execution permits.

For changed public or shared interfaces, follow contract impact. Check consequential documentation claims against the candidate, including examples, compatibility promises, and intentional internal API visibility. Use elixir-docs when applicable; separate generation evidence from executed examples and reader review.

For agent or skill changes, inspect triggers and exclusions, ownership, supported configuration, referenced resources, and representative positive/negative cases. Syntax checks do not establish correct selection or execution. Do not treat phrasing preferences as failures without a concrete effect on the reader or task.

Approval applies to the inspected candidate. Later edits require the new revision and refreshed affected review and validation; tests alone cannot approve an unreviewed rewrite.

Return one findings report to the coordinator. Deduplicate repeated findings by failure mode and affected contract/location; distinguish unresolved findings from new evidence. Do not dispatch repairs or broaden the plan. Label an interrupted or budget-exhausted review incomplete, with the coverage gap; it cannot approve the candidate.

Findings

For consequential findings include:

  • severity;
  • affected file/symbol or behavior;
  • evidence;
  • concrete failure mode;
  • suggested direction, without rewriting the production change yourself.

If no consequential issue is found, state what was reviewed and what remains unverified.

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.