quantfive/CodePress Review
Automatic code review using AI models (OpenAI, Anthropic, Google, etc.) with inline PR comments
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| github_token | GitHub token for API access | yes | ${{ github.token }} |
| model_provider | AI model provider (openai, anthropic, gemini, cohere, mistral, perplexity, fireworks, groq, xai, deepseek, openai-compatible, ollama, etc.) | yes | openai |
| model_name | AI model name or alias. Use exact names (gpt-5.2, claude-sonnet-4-5) or aliases like 'latest', 'sonnet-latest', 'gpt-mini-latest' to auto-resolve to the newest version. | yes | gpt-5.1 |
| openai_api_key | OpenAI API key (required if using OpenAI) | no | — |
| anthropic_api_key | Anthropic API key (required if using Anthropic) | no | — |
| gemini_api_key | Google Gemini API key (required if using Gemini/Google) | no | — |
| cohere_api_key | Cohere API key (required if using Cohere) | no | — |
| mistral_api_key | Mistral API key (required if using Mistral) | no | — |
| perplexity_api_key | Perplexity API key (required if using Perplexity) | no | — |
| fireworks_api_key | Fireworks API key (required if using Fireworks) | no | — |
| groq_api_key | Groq API key (required if using Groq) | no | — |
| xai_api_key | xAI API key (required if using xAI) | no | — |
| deepseek_api_key | DeepSeek API key (required if using DeepSeek) | no | — |
| openai_compatible_api_key | API key for OpenAI-compatible endpoints (required if using openai-compatible provider) | no | — |
| openai_compatible_base_url | Base URL for OpenAI-compatible endpoints (required if using openai-compatible provider) | no | — |
| ollama_api_key | Ollama API key (optional, often not needed for local Ollama instances) | no | — |
| ollama_base_url | Ollama base URL (defaults to http://localhost:11434/v1) | no | — |
| reasoning_effort | OpenAI reasoning effort level (none, minimal, low, medium, high). 'none' is only for GPT-5.1 models. | no | high |
| anthropic_effort | Anthropic effort level (low, medium, high). Only for claude-opus-4-5. Affects thinking, text responses, and function calls. | no | "" |
| thinking_enabled | Enable Anthropic extended thinking (true/false). Only for Claude claude-opus-4-5, claude-sonnet-4-5, claude-3-7-sonnet models. | no | false |
| thinking_budget | Token budget for Anthropic extended thinking (e.g., 10000, 20000). Only used when thinking_enabled is true. | no | 10000 |
| max_turns | Maximum number of turns for the interactive agent (0 = unlimited) | no | 0 |
| update_pr_description | Whether to update the PR description with AI-generated summary (true/false) | no | true |
| debug | Enable debug mode to show detailed console logs (true/false) | no | false |
| run_on_pr_opened | Run review when PR is opened (true/false) | no | true |
| run_on_pr_reopened | Run review when PR is reopened (true/false) | no | true |
| run_on_review_requested | Run review when re-review is requested from github-actions[bot] (true/false) | no | true |
| run_on_comment_trigger | Run review when comments contain trigger phrase (true/false) | no | true |
| comment_trigger_phrase | Phrase that triggers review in comments (e.g., '@codepress/review') | no | @codepress/review |
| blocking_only | Only output blocking/required comments, filter out optional/nit/fyi/praise comments (true/false) | no | false |
| force_full_review | Force a full review of all files, ignoring re-review optimizations (true/false) | no | false |
| pr_number | PR number to review (for workflow_dispatch). If not provided, will try to find PR from current branch. | no | "" |
| enable_web_search | Enable web search and fetch capabilities for the agent (true/false) | no | true |
| related_repos | YAML list of related repositories for cross-repo context. Example: related_repos: | - repo: owner/backend ref: main description: Backend API - check API contracts - repo: owner/shared ref: v2 description: Shared types and utilities | no | "" |
| related_repos_token | Token for accessing related repos (falls back to github_token for same-org repos) | no | "" |
Outputs
no outputs