gemini-review
Run a Gemini CLI code review on the current branch diff — categorizes findings as Critical/Warning/Suggestion and offers to fix critical issues
Skill by kabirdos
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/kabirdos-gemini-review
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/gemini-review
What's inside
- Skill name
gemini-review- Description
- Run a Gemini CLI code review on the current branch diff — categorizes findings as Critical/Warning/Suggestion and offers to fix critical issues
- License
- MIT
- Instructions
- 109 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# /gemini-review — Cross-Model Code Review Run Gemini as an external code reviewer on your current branch. Gets a second opinion on bugs, security, and best practices. ## Usage - `/gemini-review` — review all changes on current branch vs main - `/gemini-review --staged` — review only staged changes - `/gemini-review --file src/auth.ts` — review a specific file's changes - `/gemini-review --fix` — auto-fix critical issues without asking ## Steps ### 1. Check Prerequisites Verify `gemini` CLI is installed: ```bash command -v gemini ``` If not found, tell the user: > "Gemini CLI is not installed. Install it with: `npm install -g @anthropic-ai/gemini-cli` or check https://github.com/anthropics/gemini-cli" Stop if not available. ### 2. Get the Diff Based on flags: - Default: `git diff main...HEAD` (all changes on branch) - `--staged`: `git diff --cached` - `--file <path>`: `git diff main...HEAD -- <path>` If the diff is empty, tell the user there are no changes to review and stop. ### 3. Run Gemini Review ```bash gemini -p "You are a senior code reviewer. Review this diff for: 1. **Critical** — Bugs, security vulnerabilities, data loss risks, broken functionality 2. **Warnings** — Performance issues, missing error handling, race conditions, incomplete implementations 3. **Suggestions** — Code style, naming, simplification opportunities, best practices For each finding, include: - Category (Critical/Warning/Suggestion) - File and approximate location - What's wrong - How to fix it Be concise. Skip praise. Only report actionable findings. Diff: $(git diff main...HEAD)" ``` Capture the full output. ### 4. Parse and Display Results Show the Gemini review output, organized by severity: ``` Gemini Code Review ────────────────── Branch: feature/your-branch → main Files changed: N 🔴 Critical (N) - [file:line] Description — fix suggestion 🟡 Warnings (N) - [file:line] Description — fix suggestion 🔵 Suggestions (N) - [file:line] Description — fix suggestion ✅ No critical issues (if none found) ``` ### 5. Handle Critical Issues **If critical issues found and `--fix` flag:** - Fix each critical issue - Run tests to verify the fix doesn't break anything - Stage and commit: `fix: address Gemini review — [description]` **If critical issues found without `--fix`:** - Ask: "Found N critical issues. Want me to fix them now?" - If yes, fix, test, commit - If no, just report them **If no critical issues:** - Report the warnings and suggestions - Ask if the user wants any of them addressed ### 6. Summary ``` Review complete: N critical, N warnings, N suggestions Fixes applied: [list any commits made] ```
From kabirdos/claude-toolkit, MIT.
More from kabirdos/claude-toolkit
- 0.479score
ui-design-review
kabirdos
Acts as a UI designer — reviews a written spec/PRD or a live design (URL, local dev server, or in-repo components), audits it against proven UI design…
SkillCFreshness: CClaude Code +9 - 0.414score
insight-harness
kabirdos
Generate a comprehensive profile of your Claude Code harness — skills, hooks, workflow patterns, tool usage, token consumption, and plugin inventory across…
SkillDFreshness: DClaude Code +9 - 0.414score
manual-qa-collab
kabirdos
Collaborative manual QA testing using Playwright MCP browser automation.
SkillDFreshness: DClaude Code +9 - 0.410score
skill-showcase
kabirdos
Build a single-file public HTML showcase of a collection of Claude Code skills — table of contents, grouped categories, per-skill README with hero image…
SkillDFreshness: DClaude Code +9 - 0.409score
init
kabirdos
Bootstrap new projects with frontend design capabilities and skill creation tools.
SkillDFreshness: DClaude Code +9 - 0.409score
consensus-brainstormer
kabirdos
Use when analyzing strategic problems, evaluating options, brainstorming solutions, or making high-stakes decisions that benefit from multi-perspective…
SkillDFreshness: DClaude Code +9 - 0.409score
debate-chamber
kabirdos
Use when a problem needs iterative refinement through structured argumentation rather than independent parallel analysis.
SkillDFreshness: DClaude Code +9 - 0.409score
plan-review
kabirdos
Thorough plan review before implementation.
SkillDFreshness: DClaude Code +9 - 0.409score
qa-checklist
kabirdos
Generate an interactive QA testing checklist as a self-contained HTML page after creating a PR.
SkillDFreshness: DClaude Code +9 - 0.409score
research-orchestrator
kabirdos
Use when conducting deep research on a topic that benefits from multi-perspective analysis.
SkillDFreshness: DClaude Code - 0.406score
deploy-verify
kabirdos
Merge a PR and monitor the deployment end-to-end — watch the build, check for errors, self-heal failures, verify the deploy is clean
SkillDFreshness: DClaude Code +9 - 0.406score
bug-fix
kabirdos
Test-driven bug fixing — reproduce with a failing test first, then implement the minimal fix, iterate until all tests pass, and create a PR
SkillDFreshness: DClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 1
- Last commit (this folder)
- Mar 27, 2026 (7 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/kabirdos/claude-toolkit
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from kabirdos
- 0.479score
ui-design-review
kabirdos
Acts as a UI designer — reviews a written spec/PRD or a live design (URL, local dev server, or in-repo components), audits it against proven UI design…
SkillCFreshness: CClaude Code +9 - 0.414score
insight-harness
kabirdos
Generate a comprehensive profile of your Claude Code harness — skills, hooks, workflow patterns, tool usage, token consumption, and plugin inventory across…
SkillDFreshness: DClaude Code +9 - 0.414score
manual-qa-collab
kabirdos
Collaborative manual QA testing using Playwright MCP browser automation.
SkillDFreshness: DClaude Code +9 - 0.410score
skill-showcase
kabirdos
Build a single-file public HTML showcase of a collection of Claude Code skills — table of contents, grouped categories, per-skill README with hero image…
SkillDFreshness: DClaude Code +9 - 0.409score
consensus-brainstormer
kabirdos
Use when analyzing strategic problems, evaluating options, brainstorming solutions, or making high-stakes decisions that benefit from multi-perspective…
SkillDFreshness: DClaude Code +9 - 0.409score
init
kabirdos
Bootstrap new projects with frontend design capabilities and skill creation tools.
SkillDFreshness: DClaude Code +9
Cite this page
gemini-review skill. Agent Mart. https://agentmart.dev/skills/kabirdos-gemini-review (last updated 2026-10-24)