cassianor/agentreview
Multi-agent AI code review powered by Anthropic Claude. Posts findings as PR review comments.
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| anthropic-api-key | Anthropic API key (ANTHROPIC_API_KEY). Store as a repository secret. | yes | — |
| openai-api-key | OpenAI API key. Required only when provider is "openai". | no | "" |
| provider | LLM provider to use: "anthropic" (default) or "openai". | no | anthropic |
| agents | Comma-separated list of agents to run. Default: all agents. | no | security,performance,style,tests,docs,dependency,accessibility,i18n |
| base-ref | Git ref to diff against. Defaults to the PR base branch. | no | "" |
| model | Claude model to use (e.g. claude-sonnet-4-6, claude-opus-4-7). | no | claude-sonnet-4-6 |
| fail-on | Exit with code 1 if findings at or above this severity exist (critical|high|medium|low|never). | no | critical |
| post-comments | Post findings as inline PR review comments (requires GITHUB_TOKEN). | no | true |
| github-token | GitHub token for posting PR comments. Defaults to the automatic GITHUB_TOKEN. | no | ${{ github.token }} |
Outputs
| name | description |
|---|---|
| findings-count | Total number of unique findings after synthesis. |
| has-critical | "true" if any critical findings were found, "false" otherwise. |
| report-json | Path to the JSON report file written to the runner. |