agentleFS
Sign inStart free
agentstationagentstation/skills/autoreview/AGENTS.md

skills / autoreview

AGENTS.md1 starsChanged 2 months agoMITView on GitHub

- Upstream implementation: openclaw/agent-skills, skills/autoreview. - Record the imported upstream commit in UPSTREAM.md. - Sync upstream into a clean branch, then reapply and test the AgentStation model policy and any accepted hardening patches. - Keep repository-specific criteria and named project profiles in their owning repositories. - Invoke this helper with explicit criteria from those repositories. - Enforce the canonical sibling model-selection policy for reviewers and fallbacks. - Keep Fable coordinator-only, Astra specialist-only, and Haiku prohibited. - Check frontier eligibility before…

  • Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.

    Of the AGENTS.md files we read, the 31,843 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.

  • In this repository9 other files it gives its agents.

The file

We split this AGENTS.md into the rules it gives an agent

9 rules. About 0 are common, 7 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    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:

    Keep repository-specific criteria and named project profiles in their owning repositories.

    line 7
  2. Distinctive:

    Keep Fable coordinator-only, Astra specialist-only, and Haiku prohibited.

    line 11
  3. Distinctive:

    Check frontier eligibility before ranking candidates or installing harnesses.

    line 12
  4. Distinctive:

    Keep external adapters explicit and require concrete model identities.

    line 13
  5. Distinctive:

    Keep Opus 5.5 high as the default Claude-side code reviewer.

    line 14
  6. Very specific:

    Keep built-in Claude effort at or below high.

    line 15
  7. Distinctive:

    Keep the default automatic cadence at the substantive-code pre-PR gate.

    line 17
  8. Very specific:

    Update MODEL_SELECTION.md whenever built-in scores or benchmark inputs change.

    line 18
  9. Distinctive:

    Run the helper self-tests, Python tests, and repository skill validator after every change.

    line 20

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md 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
AGENTS.md

Autoreview maintenance

  • Upstream implementation: openclaw/agent-skills, skills/autoreview.
  • Record the imported upstream commit in UPSTREAM.md.
  • Sync upstream into a clean branch, then reapply and test the AgentStation model policy and any accepted hardening patches.
  • Keep repository-specific criteria and named project profiles in their owning repositories.
  • Invoke this helper with explicit criteria from those repositories.
  • Enforce the canonical sibling model-selection policy for reviewers and fallbacks.
  • Keep Fable coordinator-only, Astra specialist-only, and Haiku prohibited.
  • Check frontier eligibility before ranking candidates or installing harnesses.
  • Keep external adapters explicit and require concrete model identities.
  • Keep Opus 5.5 high as the default Claude-side code reviewer.
  • Keep built-in Claude effort at or below high.
  • Require conscious configuration for higher caps.
  • Keep the default automatic cadence at the substantive-code pre-PR gate.
  • Update MODEL_SELECTION.md whenever built-in scores or benchmark inputs change.
  • Run the helper self-tests, Python tests, and repository skill validator after every change.

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.

Reviews can't be read right now.

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.

Check your setup, free