agentleFS
fqmyysjjdfqmyysjjd/agent-feed/.agents/skills/project-review/SKILL.md

project-review

Skill4 starsChanged 6 months agoMITView on GitHub

Use when reviewing modified code, git diffs, commits, merges, PR-like changes, or implementation output.

17 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 12 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:

    Use this skill for code, diff, commit, merge, or AI-generated implementation review.

    line 12
  2. Distinctive:

    Do not modify code during a review unless the user explicitly asks for fixes in the same task.

    line 14
  3. Very specific:

    If a finding requires an unconfirmed decision outside the current Task Brief, stop and apply .agents/rules/decision-gates.md.

    line 25
  4. Very specific:

    Use .agents/skills/concept-review/SKILL.md when the diff introduces or changes names, concepts, abstractions, protocol terms, user-facing …

    line 45
  5. Distinctive:

    Check for evidence of:

    line 51
  6. Distinctive:

    If a relevant item has no evidence, report it as a finding or residual risk.

    line 64

“Common” is measured: kept by at least 6 in 10 of the 12 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

Project Review Skill

Required Use

Use this skill for code, diff, commit, merge, or AI-generated implementation review.

Do not modify code during a review unless the user explicitly asks for fixes in the same task.

Review Mode

Classify the review before acting:

  1. Pure review: the user asked for review, diff review, commit review, or findings. Do not edit files. Return findings and residual risk only.
  2. Implementation gate: the review is part of a task that already includes coding, refactor, fix, or tests. Findings may become follow-up fixes inside the same Task Brief.

For implementation gates, route P0/P1 findings and default-fix P2 findings through .agents/skills/project-fix/SKILL.md, then rerun relevant verification before final handoff.

If a finding requires an unconfirmed decision outside the current Task Brief, stop and apply .agents/rules/decision-gates.md.

Required Reading

  1. .agents/rules/outcome-boundary.md
  2. .agents/project/architecture-boundaries.md
  3. .agents/rules/development-workflow.md
  4. .agents/project/project-structure.md
  5. .agents/rules/testing-gates.md
  6. .agents/rules/review-gates.md
  7. .agents/domain/contracts.md
  8. .agents/skills/README.md
  9. .agents/skills/specialist-router/SKILL.md when an optional, imported, custom, or specialized skill may match the diff or risk.

Review Scope

Findings should prioritize correctness, contract drift, module ownership, tests, error handling, security/secret handling, and maintainability.

Before finalizing findings, check .agents/skills/README.md for specialized review or fix skills that match the changed language, architecture, framework, persistence layer, or risk category. Use .agents/skills/specialist-router/SKILL.md when the match is not obvious, when a custom skill is involved, or when multiple skills could apply. Use only the skills that directly serve the current review boundary.

Use .agents/skills/concept-review/SKILL.md when the diff introduces or changes names, concepts, abstractions, protocol terms, user-facing language, or skill vocabulary.

Module Understanding Review

For every affected module, command, adapter, protocol asset, or public surface, review whether the change proves the AI understood the path it modified.

Check for evidence of:

  1. Affected entrypoints, owners, and outputs: the changed files cover the actual requirement path and do not miss a relevant caller, adapter, command, generated template, or user-visible surface.
  2. Flow closure: input or trigger -> orchestration/state -> output, write, generated asset, or side effect is complete.
  3. Contract consistency: function signatures, CLI arguments, config keys, schemas, generated file paths, adapter expectations, and documented behavior stay aligned.
  4. Boundary and failure handling: validation, permissions, error messages, cleanup, rollback expectations, and logging/audit anchors are handled at the owning layer.
  5. State and data safety: defaults, empty/null cases, async or repeated execution, durable state, secrets, and trust/config boundaries remain safe.
  6. Reuse and duplication: existing helpers, patterns, tests, adapters, or template owners were reused before adding parallel behavior.
  7. Maintainability cost: new concepts, abstractions, branches, or files add only necessary complexity.
  8. Minimal implementation: the diff does not introduce future-facing machinery, broad rewrites, or unrelated formatting/noise to satisfy a narrow task.
  9. Style consistency: similar existing behavior uses compatible naming, layering, errors, output style, comments, and verification patterns.
  10. Shared or deletion risk: global files, shared helpers, generated assets, removed paths, and adapter mirrors were checked for other consumers.

If a relevant item has no evidence, report it as a finding or residual risk. Do not replace missing evidence with a general "looks fine" statement.

Severity Model

  1. P0: Data loss, secret leak, source-of-truth break, or irreversible corruption.
  2. P1: Public API, contract, lifecycle, persistence, security, or recovery break.
  3. P2: Module boundary, test coverage, traceability, or maintainability risk.
  4. P3: Naming, local readability, or small cleanup issue.

Output

Return findings first, ordered by severity. If no findings are found, say so explicitly and list residual risks or tests not run.

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.