azogoat/AI Reviewer
Ask for a review on a PR and get a normal GitHub review back, inline comments and all
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| model | Model id string, claude-* (Anthropic) or gpt-*/o* (OpenAI) | — | claude-opus-4-8 |
| max_tool_calls | Cap on exploration tool calls; a runaway guard, not a depth limit | — | 50 |
| exploration_token_budget | Optional cap on total tokens across exploration steps | — | "" |
| max_inline_comments | Maximum inline review comments; the rest go into the review body | — | 15 |
| inline_severity_threshold | Minimum severity posted inline (critical, major, minor); nits never post inline | — | minor |
| request_changes_threshold | Minimum severity that turns the verdict into REQUEST_CHANGES | — | major |
| reviewer_login | Machine user login that must be the requested reviewer in team mode | — | light-ai-reviewer |
| pull_number | PR number, only needed for workflow_dispatch runs | — | "" |
| dry_run | Log the would-be review instead of posting it | — | false |
| github_token | Token used to read the PR and submit the review | yes | — |
| anthropic_api_key | Anthropic API key, required for claude-* models | — | "" |
| openai_api_key | OpenAI API key, required for gpt-*/o* models | — | "" |
Outputs
no outputs