samushi/AI Code Review Agent
AI-powered, auto-stack code reviews for PRs (React/Next, Vue/Nuxt, Laravel/WordPress, Django/Flask …)
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| github-token | GitHub token with repo scope (PR comments) | yes | ${{ github.token }} |
| provider | LLM provider: openai | anthropic | gemini | ollama | no | openai |
| openai-api-key | OpenAI API key (if provider = openai) | no | — |
| anthropic-api-key | Anthropic Claude API key (if provider = anthropic) | no | — |
| gemini-api-key | Google Generative AI key (if provider = gemini) | no | — |
| ollama-base-url | Base URL of your Ollama server (if provider = ollama) | no | http://localhost:11434 |
| ai-model | Model name for the chosen provider | no | gpt-4o-mini |
| stack | Force stack (react, vue, laravel, django, …) or leave blank for auto-detect | no | — |
| file-patterns | Comma-separated glob patterns to include | no | **/*.{js,jsx,ts,tsx,vue,php,blade.php,py,json,yml} |
| exclude-patterns | Comma-separated glob patterns to ignore | no | **/{node_modules,dist}/** |
| min-score-threshold | Minimum score before failing CI (with fail-on-low-score = true) | no | 7 |
| post-comment | Whether to post the AI report as a PR comment | no | true |
| fail-on-low-score | Fail the job when review-score < min-score-threshold | no | false |
Outputs
| name | description |
|---|---|
| review-score | Numeric score (1-10) |
| recommendation | POSITIVE | NEEDS_CHANGES | NEGATIVE |
| issues-found | Number of findings reported by the AI |
| review-summary | One-paragraph summary of the review |