eljonathas/GH Review Agent

Run the shared PR review engine from a GitHub Actions workflow

View on GitHub

Trust Signals

Scorecard Score
not yet scored
Maintenance Recency
Stale
License
None
namedescriptionrequireddefault
github_tokenGitHub token used to read the PR and publish the reviewyes
llm_providerLLM provider to use: openai or openrouternoopenai
llm_api_keyAPI key for the configured LLM provideryes
llm_base_urlOptional base URL override for the configured LLM providerno
ownerGitHub repository owner overrideno
repoGitHub repository name overrideno
pull_numberPull request number overrideno
llm_modelLegacy fallback model for the main review loopno
llm_model_mainModel used by the main review loop that decides tool callsno
llm_model_taskLightweight model used for auxiliary tasks such as repair/summariesno
llm_model_reviewLegacy alias for llm_model_mainno
llm_model_explorerLegacy alias for llm_model_taskno
llm_model_narratorModel used for the narrator phaseno
max_turns_per_fileOptional override for the per-file review loop limitno
max_turns_explorerOptional override for the explorer loop limitno
review_concurrencyOptional override for parallel file review concurrencyno
budget_tokensTotal token budget shared across all agents in a review session. When set, agents self-regulate depth based on remaining budget.no

no outputs