- Home
- Skills
- Google Gemini
- Gemini Cli
- Code Reviewer
code-reviewer_skill
- TypeScript
- Official
92.6k
GitHub Stars
1
Bundled Files
3 weeks ago
Catalog Refreshed
2 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstart where the catalogue uses aiagentskills.
npx veilstart add skill google-gemini/gemini-cli --skill code-reviewer- SKILL.md3.1 KB
Overview
This skill reviews code changes for both local development and remote Pull Requests. It focuses on correctness, maintainability, readability, efficiency, security, edge cases, and testability. The goal is actionable feedback framed as summary, findings (critical, improvements, nitpicks), and a clear recommendation.
How this skill works
For remote PRs, the skill checks out the PR (via GitHub CLI), runs the project's preflight suite, and reads the PR description and comments for context. For local changes, it inspects the working tree and staged diffs, optionally prompting to run preflight checks. It then analyzes changes against core review pillars and returns a structured report with suggested fixes and test recommendations.
When to use it
- Review a remote Pull Request by number or URL.
- Review local staged or unstaged changes before committing.
- Get a focused audit for correctness, security, or performance concerns.
- Prepare feedback before merging a feature branch.
- Validate that tests and edge-case handling are sufficient.
Best practices
- Provide the PR number or URL to target a remote review explicitly.
- Run the project’s preflight/test suite when reviewing substantial changes.
- Share any project-specific style or architecture guidelines up front.
- Limit reviews to a single concern set (e.g., bugfix, refactor) per run for clarity.
- Ask for missing context (design decisions, intended behavior) before making major suggestions.
Example use cases
- "Review PR #123" — checkout PR, run preflight, and produce a structured review.
- "Review my working changes" — inspect git diff and staged files, then report issues to fix before commit.
- Security check — focus the review on potential vulnerabilities and insecure patterns.
- Performance audit — identify obvious bottlenecks and suggest optimizations.
- Test coverage recommendations — propose unit or integration tests for uncovered edge cases.
FAQ
Yes. By default the skill inspects the working tree and staged changes; it can focus only on staged files if requested.
Do you run tests automatically for remote PRs?
For remote PRs the skill runs the project’s standard preflight/test script (e.g., npm run preflight) to surface automated failures before manual review.