eljonathas/GH Review Agent
Run the shared PR review engine from a GitHub Actions workflow
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| github_token | GitHub token used to read the PR and publish the review | yes | — |
| llm_provider | LLM provider to use: openai or openrouter | no | openai |
| llm_api_key | API key for the configured LLM provider | yes | — |
| llm_base_url | Optional base URL override for the configured LLM provider | no | — |
| owner | GitHub repository owner override | no | — |
| repo | GitHub repository name override | no | — |
| pull_number | Pull request number override | no | — |
| llm_model | Legacy fallback model for the main review loop | no | — |
| llm_model_main | Model used by the main review loop that decides tool calls | no | — |
| llm_model_task | Lightweight model used for auxiliary tasks such as repair/summaries | no | — |
| llm_model_review | Legacy alias for llm_model_main | no | — |
| llm_model_explorer | Legacy alias for llm_model_task | no | — |
| llm_model_narrator | Model used for the narrator phase | no | — |
| max_turns_per_file | Optional override for the per-file review loop limit | no | — |
| max_turns_explorer | Optional override for the explorer loop limit | no | — |
| review_concurrency | Optional override for parallel file review concurrency | no | — |
| budget_tokens | Total token budget shared across all agents in a review session. When set, agents self-regulate depth based on remaining budget. | no | — |
Outputs
no outputs