agentleFS
odytriceodytrice/metaskills/skills/code-review/SKILL.md

code-review

Skill1 starsChanged 14 days agoMITView on GitHub

Review PRs for merge or local changes for commit readiness. Use when asked to review a PR, branch, diff, or current changes.

194 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use close-spec, design-with-images and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • dzhngSkill

    close-spec

    When a spec under specs/<feature>/ is done shipping (a write-spec build, or any planned task that produced a spec), archive it to specs/done/ and rewrite it from a build-plan into a durable rationale — the why, the principles, the invariants — pointing back to the real code for the how. Use when implementation has landed and the plan no longer matches what shipped, or the user says a feature/spec is finished. Pairs with [write-spec](../write-spec/SKILL.md) (the plan this closes) and [review](../review/SKILL.md) (run before closing).

    Also in 5 projects that use this

  • dzhngSkill

    design-with-images

    Explore UI and visual design with image generation, then iterate the real implementation against the user-selected concept. Use when brainstorming visual directions, offering mockup variants, or implementing an approved generated design; do not stop at the mockups or the first code approximation.

    Also in 5 projects that use this

  • PydanticSkill

    add-new-model

    Add support for a newly-released language or image generation model in pydantic-ai (e.g. openai:gpt-5.6, anthropic:claude-sonnet-5, openai:gpt-image-2). Use when a provider ships a new model id and you need to wire literals, profile flags, adapters, and tests to recognize it. Handles SDK-lag, gateway list conventions, capability probing, and direct image-model geometry.

    Also in 9 projects that use this

  • transitive-bullshitSkill

    compare-screenshots

    Compare screenshots against the intended design, distinguishing approved references from historical baselines. Use for iterative implementation matching, before/after visual review, objective image telemetry, or checking a lone capture for flat, empty, or badly framed content.

    Also in 5 projects that use this

  • Microsoft AzureSkill

    azsdk-common-apiview-feedback-resolution

    Analyze and resolve APIView review feedback on Azure SDK PRs. **UTILITY SKILL**. USE FOR: APIView comments, API review feedback, SDK API surface changes. DO NOT USE FOR: general code review, non-APIView feedback. INVOKES: azure-sdk-mcp:azsdk_apiview_get_comments, azure-sdk-mcp:azsdk_customized_code_update, azure-sdk-mcp:azsdk_typespec_delegate_apiview_feedback, azure-sdk-mcp:azsdk_run_typespec_validation, azure-sdk-mcp:azsdk_package_generate_code.

    Also in 6 projects that use this

  • Microsoft AzureSkill

    azsdk-common-generate-sdk-locally

    Generate and finalize Azure SDKs locally from TypeSpec, including build, checks, tests, changelog, metadata, and version updates by default. WHEN: "generate SDK locally", "build SDK", "run SDK tests", "run CI checks", "validate package", "run checks", "update changelog", "fix SDK build errors", "resolve SDK generation errors", "customize TypeSpec", "rename SDK client", "rename SDK model", "hide operation from SDK", "fix analyzer errors", "resolve customization drift", "create subclient", "update metadata", "update version". INVOKES: azsdk_verify_setup, azsdk_package_generate_code, azsdk_package_build_code, azsdk_package_run_check, azsdk_package_run_tests, azsdk_customized_code_update, azsdk_package_update_changelog_content, azsdk_package_update_metadata, azsdk_package_update_version.

    Also in 6 projects that use this

  • +4 more from dzhng
  • +2 more from Azure

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 184 edited versions we read keep these.

  • 2

    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. Very specific:

    Use this directory's checklist.md plus § Review Notes.

    line 10
  2. Very specific:

    Remove only a clean worktree whose resolved path is inside the repo's .worktrees/.

    line 41
  3. Distinctive:

    Read changed files and surrounding context via § Code Layout & Tech Stack.

    line 46
  4. Distinctive:

    No worktree or separate agent.

    line 78

“Common” is measured: kept by at least 6 in 10 of the 184 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

Code Review

Select mode first. Both require AGENTS.md § Build & Validation (commands, DB tripwires, commit convention), § Code Layout & Tech Stack, § Review Notes (checks/detail file). PR additionally requires § Repositories, § Branch Map (bases, auto-deploy), § Project Board (done automation). Missing required section: name it and stop.

Use this directory's checklist.md plus § Review Notes; output-format.md defines output/severity. Findings first; style only where it hides risk.

PR: explicit board: none skips board operations per issue-plan § Board Status Transitions; missing facts stop. Local: no board config/operations. Standalone partial PRs are allowed; transition only issues this PR actually closes, never mere references.

Modes

  • PR mode (PR number/URL/branch): independent agent/worktree, PR comment, conditional squash-merge.
  • Local mode (no target, uncommitted changes, pre-commit gate): in-place review, user report; any Critical/High blocks commit.

PR Mode

Reviewer must differ from implementer. If you authored/substantially edited the change, stop and spawn a reviewer with only PR number/URL, linked issue, validation expectations; never reuse implementation context/worktree. Apply dev-cycle § Subagent Dispatch: foreground by default, background only on explicit user request. Reviewer alone merges; caller (user, dev-cycle, burndown) only records results.

Review Worktree

Capture <reviewed-sha> = headRefOid, <base-sha> = baseRefOid, <base> = baseRefName; cross-check base against § Branch Map. Verify origin matches § Repositories' PR base repo; scope number-based gh commands there. Fetch PR head ref (also for forks); require captured SHA match before detached checkout:

git fetch origin <base>
git fetch origin refs/pull/<number>/head
git rev-parse FETCH_HEAD
git worktree add --detach .worktrees/review-pr-<number> <reviewed-sha>

Require worktree git rev-parse HEAD = <reviewed-sha> and both captured commits available. All reads/diffs/validation run here using git diff <base-sha>...<reviewed-sha>, never moving gh pr diff. Fetch/head mismatch: fresh metadata and review, never substitute newer SHA. After comment/any merge, from original repo:

git worktree remove .worktrees/review-pr-<number>
git worktree prune

Remove only a clean worktree whose resolved path is inside the repo's .worktrees/; otherwise stop/report.

Process

  1. gh pr view <number> --json title,body,baseRefName,baseRefOid,headRefName,headRefOid,url,closingIssuesReferences; capture SHAs before checkout. Read pinned-range paths/commits, linked issues, and paginated REST ledger comments per issue-plan; ledger = intended design.
  2. Read changed files and surrounding context via § Code Layout & Tech Stack; apply relevant checklist/Review Notes.
  3. Validate touched layers where practical per § Build & Validation. Touched DB tripwire: live-database suite must pass locally or in CI before approval.
  4. Post per output-format.md, header **Review complete** -- [View PR](<pr-url>) (use View job if job URL exists), naming reviewed head/base SHAs. Re-fetch metadata; changed head requires fresh pinned checkout, validation, re-review before approval/merge. Never transfer approval to a new SHA.

Re-review (revision passes)

Existing **Review complete** comment: converge.

  • Classify each prior finding Resolved, Partially resolved, or Not resolved, citing commit/hunk.
  • New findings only: fix-induced regressions or release-threatening Critical/High not surfacable earlier. No previously reviewable new Medium/Low.
  • Open with Prior findings recap; list only open findings/new regressions; label header re-review.
  • dev-cycle bounds revisions: two, then park.

Squash Merge (the reviewer merges)

Merge only if Approve, no Critical/High or important pre-merge Medium, required checks passed, PR mergeable, no production-breaking validation, no review-only request, and no § Branch Map auto-deploy the user wants to time manually.

gh pr merge <number> --repo <repo-slug> --squash --match-head-commit <reviewed-sha>

Immediately before merge re-fetch head/base and reviewed-head required checks. Changed head/base or head-match rejection: refresh pinned review/validation, never just retry a newer SHA. Use PR commit subject if it meets § Build & Validation convention; otherwise write one that does. Unmet criterion: leave open and report it.

After A Merge

  1. Board. board: none: skip/report. Otherwise verify registered closing issues are closed. Automation off: move only closed issues to done per issue-plan § Board Status Transitions. Automation on: verify with the same complete lookup; report pending/failure honestly. Last open sub-issue: note parent closeable, never close it.
  2. Branches. Detached checkout creates no local head branch. Preserve pre-existing user/implementation branches; never use --delete-branch. Remote cleanup belongs to repository automation or a separate explicit request.

Report: comment URL, assessment, merge commit or blocking criterion, board transition, worktree and branch cleanup.

Local Mode

No worktree or separate agent.

  1. git diff --name-only and git diff --cached --name-only (or caller's paths). No changes: report/stop.
  2. Map files via § Code Layout & Tech Stack; read diffs/surroundings and apply relevant checklist/Review Notes.
  3. Validate where practical; touched DB tripwire requires live-database suite before commit.
  4. Report per output-format.md. Any Critical/High blocks commit: Summary says "Commit blocked: <n> Critical/High" and lists required fixes. Re-review with the same convergence discipline.

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.