
Gemini
- Updated May 16, 2026
- ufaf0052/gemini-plugin-cc
gemini is a Claude Code skill in the AI & Agent Building category. Use Gemini from Claude Code to review code or delegate tasks.
Key points
- gemini
- AI & Agent Building
- AI-coding skill
Gemini by the numbers
- Data as of Jul 7, 2026 (Skillselion catalog sync)
/plugin marketplace add ufaf0052/gemini-plugin-cc/plugin install gemini@gemini-plugin-ccAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Last updated | May 16, 2026 |
|---|---|
| Repository | ufaf0052/gemini-plugin-cc ↗ |
What it does
Use Gemini from Claude Code to review code or delegate tasks.
README.md
Gemini plugin for Claude Code
Use Gemini from inside Claude Code for code reviews or to delegate tasks to Gemini.
This plugin is for Claude Code users who want an easy way to use Gemini's adversarial code review from the workflow they already have.
What You Get
/gemini:reviewfor an adversarial read-only Gemini review/gemini:adversarial-reviewfor a steerable challenge review with custom focus text/gemini:rescue,/gemini:status,/gemini:result, and/gemini:cancelto delegate work and manage background jobs/gemini:setupto check Gemini CLI readiness and manage the stop-time review gate
Requirements
- Gemini CLI installed and accessible (via
GEMINI_BINenv or at/opt/node/bin/gemini) - Node.js 18.18 or later
Install
Add the marketplace in Claude Code:
/plugin marketplace add ufaf0052/gemini-plugin-cc
Install the plugin:
/plugin install gemini@gemini-plugin-cc
Reload plugins:
/reload-plugins
Then run:
/gemini:setup
/gemini:setup will tell you whether Gemini is ready. If Gemini CLI is not found, it will provide guidance on setting the GEMINI_BIN environment variable.
After install, you should see:
- the slash commands listed below
- the
gemini:gemini-rescuesubagent in/agents
One simple first run is:
/gemini:review --wait
/gemini:status
/gemini:result
Usage
/gemini:review
Runs an adversarial Gemini review on your current work. Returns structured JSON output with verdict, findings (severity + file:line + confidence), and next steps.
Use it when you want:
- a review of your current uncommitted changes
- a review of your branch compared to a base branch like
main
Use --base <ref> for branch review. Supports --wait and --background.
Examples:
/gemini:review
/gemini:review --base main
/gemini:review --background
/gemini:adversarial-review
Same as /gemini:review but accepts custom focus text to steer the adversarial review toward specific concerns.
Examples:
/gemini:adversarial-review focus on auth and API contract changes
/gemini:adversarial-review --base main focus on race conditions
/gemini:rescue
Delegate investigation, diagnosis, or analysis tasks to Gemini.
Examples:
/gemini:rescue diagnose why the auth tests are failing
/gemini:rescue --background analyze the performance regression in the API layer
/gemini:setup
Check Gemini CLI availability and manage the stop-time review gate.
/gemini:setup
/gemini:setup --enable-review-gate
/gemini:setup --disable-review-gate
/gemini:status
Show active and recent Gemini jobs.
/gemini:status
/gemini:status <job-id>
/gemini:result
Retrieve stored final output for a finished Gemini job.
/gemini:result
/gemini:result <job-id>
/gemini:cancel
Cancel an active background Gemini job.
/gemini:cancel
/gemini:cancel <job-id>
Architecture
This plugin is modeled after openai/codex-plugin-cc. Key differences:
- No app-server/broker — Gemini CLI is invoked via
spawnwith stdin pipe (stateless) - No persistent sessions — no thread resume or session management
- Adversarial by default — both
/gemini:reviewand/gemini:adversarial-reviewuse the same XML-structured adversarial prompt with grounding rules, calibration rules, and structured JSON output
Environment Variables
| Variable | Default | Description |
|---|---|---|
GEMINI_BIN |
/opt/node/bin/gemini |
Path to Gemini CLI binary |
GEMINI_MODEL |
gemini-3-flash-preview |
Model to use |
GEMINI_TIMEOUT_MS |
600000 (10 min) |
CLI execution timeout |
Important: GEMINI_MODEL in shell profile (e.g. ~/.bashrc) takes precedence over the code default. If reviews are unexpectedly slow or timing out, check echo $GEMINI_MODEL first.
Troubleshooting
Gemini review hangs or times out in pipelines
Symptom: Reviews work in interactive terminal but hang when run from claude -p or CI pipelines.
Root cause: Gemini CLI Issue #6715 — the CLI hangs when spawned as a subprocess via child_process.
Fix (applied in v1.0.1+): runGeminiPrompt() uses spawn() + stdin pipe instead of execFile() + -p argument. The prompt is written to stdin; -p "" triggers non-interactive mode.
Wrong model being used
Symptom: ps aux | grep gemini shows a different model than configured in code.
Root cause: GEMINI_MODEL environment variable in ~/.bashrc or ~/.profile overrides the code default.
Fix: Update or remove the env var:
# Check current value
echo $GEMINI_MODEL
# Update in ~/.bashrc
export GEMINI_MODEL=gemini-3-flash-preview
Partial output on timeout
When a review times out, the plugin now returns any partial stdout collected before the kill signal, instead of discarding it. This means slow reviews may still produce usable findings.
License
Apache-2.0