Code-Review plugins

Reusable review assets for internal developer-platform repositories: API style and security skills, a service-review and a test-generation command, and a security-reviewer subagent.

4 skills · 2 commands · 1 agent

Six companion skills: codereview-inline (process @AI comments), codereview-clean (sweep the trail), commit-reply-review (snapshot dirty tree before the reply pass), diffity-resolve-traced and diffity-resolve-tree-traced (wrap nilbuild/diffity resolution flows and leave an @SEEN @AI:/@AI-reply: trail at every edit site), diffity-commit-resolve-traced (snapshot dirty tree before the traced diffity resolve pass).

6 skills

Personal Claude Code skills: ISO-25010 quality review, coding practice review, i18n violation fix, and test authoring/extraction

5 skills

Turn code changes into an interactive, animated local walkthrough with reviewer feedback that flows back to the agent.

1 skill · 4 commands

AI code review with specialized sub-agents for logic, security, and edge cases

1 command · 3 agents · hooks

Publish review manifests with your pull requests, run an independent security review, and review agent-authored PRs by what matters.

3 skills · 1 agent · hooks

Karpathy's 4 coding principles + 10-section production audit. Service-profile-driven, grade-aware, multilingual (EN/한국어, responds in any language).

2 skills · 2 commands

Severity-ranked code review workflows covering general correctness, security, performance, and pull request review against stated requirements.

4 skills

빌드 도구 없는 바닐라 웹(단일 HTML, 인라인 JS/CSS, Canvas, localStorage) 전용 코드 품질 키트. XSS·저장소 예외·이벤트 누수·애니메이션 경합·무작위 편향·경계값·접근성을 훑는 web-review 스킬, 그 스킬로 리뷰를 수행하고 코드는 고치지 않는 web-reviewer 서브에이전트, 커밋 전에 검사 스크립트를 강제하는 PreToolUse 훅으로 이루어져 있다.

1 skill · 1 agent · hooks

Maintainer PR review queue: finds the PRs waiting on you, reviews each one in an isolated worktree with a read-only subagent, and drafts pending GitHub reviews that only you can see.

1 skill · 1 agent · hooks

Reviews OSRS plugin, app and extension code for API etiquette and Plugin Hub readiness: per-item request loops, blocked User-Agents, missing caching, polling faster than the data changes, expensive parser queries where a bulk dump exists, and features gated behind payment on non-commercially-licensed data.

1 skill · 2 commands

Code review, linting guidance, and quality enforcement skills

1 skill · 1 agent

Enforce team PR review conventions: changelog entry, no console.log, test coverage, no formatting churn, security checklist.

1 skill · 1 command

AI-driven PR review — `pr-review` skill for user-invoked review of the current branch, `pr-review` subagent for programmatic dispatch by other agents (e.g. ship). Both wrap the agent-tool-pr-reviewer CLI.

1 skill · 1 agent

/review:analyze command — triage code review feedback into FIX / IGNORE / DISCUSS with reasoning.

1 command

Audits skills repos

1 skill

PR を sanity-review 流儀（実測検証 + Codex 相談 + .dev/sanity-review 保存 + PR コメント投稿）でレビューする。shokai の sanity-review スキルを実運用向けに具体化した手順。

1 skill

Interactive code review workflow with per-finding decisions

1 skill

Address external review findings through a watcher -> implementer -> verifier loop with model-pinned subagents. Implementers never commit; an independent verifier attacks the merged result by execution before anything ships.

1 skill

Multi-agent end-to-end backend/API PR review: phased orchestrator with parallel review passes, real-request runtime verification against a project-configured non-prod target, and a PR-ready verdict. Auto-configures a per-repo review extension.

1 skill

Interactive code review workflow with per-finding decisions

1 skill

Review of what would hurt a software system under stress or through a mistake: traces one critical flow, rates the control that stops each human or coding-agent mistake, scores exposure, and either reports findings or patches the highest-risk reversible issue. Bundles a dependency-free scanner, an evidence checker, and a test-strength checker.

1 skill

Read-only architecture reviewer that checks a diff against the project's documented architecture contracts

1 agent

Enforce team conventions on every pull request: changelog, tests, security checklist, no churn.

1 skill

Prove review findings against the code and fix only the valid ones; `converge` repeats failing test, fix and independent review for up to three rounds, and a Stop hook keeps the loop from ending early

1 skill

Code review, compare, and debate tools using multiple AI models

Codex reviews as background jobs with findings to pick and send to Claude, and a CodexReview tool