keploy/Keploy-AI Code Review
Provides automated, AI-powered code reviews on pull requests with a resilient fallback system using GitHub Models and Ollama.
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| github-token | Required. The GITHUB_TOKEN secret. | yes | — |
| model | The primary AI model to use (e.g., from GitHub Models). | no | gpt-4o-mini |
| ollama-model | The Ollama model to use as a fallback if the primary model fails. | no | qwen2:7b |
| use-ollama-fallback | Enable the Ollama fallback system. | no | true |
| max-tokens | The token limit for smart diff prioritization. Diffs larger than this will be summarized. | no | 8000 |
| temperature | The creativity/randomness of the AI model (0.0 to 2.0). | no | 0.2 |
| ignore-patterns | Comma-separated list of file patterns to ignore (e.g., "*.md,dist/**"). | no | "" |
| include-patterns | Comma-separated list of file patterns to force-include (overrides ignore). | no | "" |
Outputs
| name | description |
|---|---|
| review-posted | "true" if a review comment was successfully posted, otherwise "false". |
| review-provider | The provider that generated the review (github-models, ollama, or static-fallback). |