yamcodes/arkenv/code-review
Claude Code and agent skill by yamcodes/arkenv
- community
Checking that the SKILL.md file is still in the repository.
Fetch, analyze, and address code reviews and comments on GitHub, or perform a code review on changes in the workspace. Make sure to use this skill whenever the user mentions code reviews, addressing PR reviews, fixing PR comments, resolving review feedback, performing a code review, or running '/code-review', even if they don't explicitly name GitHub or the PR number.
Install
npx vskill@latest install yamcodes/arkenv/code-reviewSKILL.md
Code Review
A dual-purpose skill to either (1) systematically retrieve and address pull request review comments on GitHub, or (2) perform a code review on changes in the workspace.
Workflow Selection
At the start of execution, identify the user's intent:
- Mode A: Address / Fix Review Comments: Follow this mode if the user wants to address, fix, or resolve comments/feedback from a prior or active review (e.g., "fix PR comments", "address review notes", "resolve feedback", "address /code-review").
- Mode B: Perform Code Review: Follow this mode if the user wants a review of their changes or the current workspace (e.g., "review my code", "check these changes", "do a code review", "/code-review").
Mode A: Address / Fix Review Comments
Systematically retrieve and fix comments from a PR review.
1. Locate current PR and comments
- Run
git statusto find the current branch. - Run
gh pr statusto find the active pull request number and repository information. - Fetch Comments (Avoid Missing Inline Comments): Standard commands like
gh pr view <PR_NUMBER> --commentsonly return high-level conversation summaries and completely omit inline code review comments made in the diff. To reliably retrieve the complete list of inline review comments, query the GitHub API directly:
``bash gh api repos/:owner/:repo/pulls/<PR_NUMBER>/comments --jq '.[] | {id, path, line, body}' ``
2. Parse and analyze comments
- Extract the list of review comments. Focus on unresolved or actionable requests.
- For each comment, identify the affected file path, line numbers, and expected outcome.
- Check for Stale/Addressed Comments: Before making any code modifications, verify if the comment is already addressed or stale (e.g. if the target code has already been changed or deleted in a recent commit). Mark such comments as "Stale/Already Addressed" and skip them.
- Handle Ambiguity / Ask for Clarification: If a comment is ambiguous, lacks context, or if you are unsure of the best implementation path, STOP and ask the user for clarification. Do not make risky assumptions.
3. Implement the fixes
- View the surrounding code using the
view_filetool to understand context. - Update the code using
replace_file_contentormulti_replace_file_content. Keep changes minimal and adhere strictly to the codebase's existing code style, imports, and naming patterns.
4. Verify changes
- Compile/build the package (e.g.,
pnpm build,npm run build, ortsc --noEmit). - Run the test suite (e.g.,
pnpm test,npm test, orvitest) to ensure all tests pass. - Run size limit checks if applicable.
5. Report results & comment back
- Present a clear markdown table summarizing:
- Review comment / concern.
- File and lines changed.
- Action taken (e.g., "Fixed", "Already Addressed", "Skipped").
- Verification status.
- Comment and Resolve on PR: Always reply to and resolve addressed threads on GitHub:
- Reply: Post a reply to the thread using the REST API (with the database ID of the last comment):
``bash gh api -X POST /repos/:owner/:repo/pulls/<PR_NUMBER>/comments/<comment_id>/replies -f body="Addressed in latest commit." ``
- Resolve Thread: Mark the thread as resolved using the GraphQL API (using the GraphQL thread ID
PRRT_...):
``bash gh api graphql -F threadId="<thread_id>" -f query=' mutation($threadId: ID!) { resolveReviewThread(input: { threadId: $threadId }) { thread { id isResolved } } } ' ``
- General Comment: To add a general comment to the PR:
``bash gh pr comment <PR_NUMBER> --body "Addressed review comments: ..." ``
Mode B: Perform Code Review
Perform a high-quality, professional code review of the workspace changes.
1. Retrieve changes
- Run
git statusto check modified files. - Run
git diff(or compare branches) to retrieve the full diff of the changes under review.
2. Analyze the changes
Review the diff/code for:
- Correctness & Bugs: Off-by-one errors, edge cases, incorrect conditional logic.
- TypeScript & Types: Stale or weak types, unsafe casts, unused imports or variables.
- Dead Code: Functions, imports, or files that are defined but never used.
- Stale JSDoc & Docs: Stale doc comments that refer to renamed functions or old signatures.
- Architectural Patterns: Consistency with the repository's architecture (e.g. modular structure, proper abstraction boundaries).
3. Report review findings
Output the code review structured by file:
- For each file, list specific findings with line numbers and recommendations.
- Categorize findings into:
- ⚠️ Critical / Issues: Bugs, type/compilation errors, or logic issues.
- ℹ️ Refactors / Cleanup: Dead code, unused imports, stale JSDoc.
- 💡 Suggestions / Nits: Minor styling, readability improvements, or alternative approaches.
Security
- Security scanPassed, 96 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/code-review/SKILL.md
Popularity
- Repository stars
- 144
- Repository forks
- 6
- 7-day trend score
- 30
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/yamcodes/arkenv/code-review)