mariorazo97/AI CI Healer
🚑 Stops builds from failing. Analyzes logs & auto-fixes errors using AI (Groq/Gemini/Ollama).
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| github-token | GitHub Token to post comments (use secrets.GITHUB_TOKEN) | yes | — |
| llm-provider | AI Brain to use: "groq" (fastest), "gemini" (smartest), or "ollama" (private) | no | groq |
| model | Specific model to use (e.g., "llama-3.1-70b-versatile" or "gemini-1.5-flash") | no | "" |
| groq-api-key | Groq API Key (Required if provider is groq) | no | — |
| gemini-api-key | Gemini API Key (Required if provider is gemini) | no | — |
| ollama-endpoint | Ollama Endpoint for self-hosted execution (e.g., http://localhost:11434) | no | — |
| custom-context | Teach the AI your coding standards (e.g., "Always use TypeScript strict mode") | no | "" |
| max-log-lines | How much log history to read (Increase for complex errors) | no | 500 |
| confidence-threshold | Minimum confidence % required to post a comment | no | 50 |
| enable-comments | Post the fix as a comment on the PR? (true/false) | no | true |
Outputs
| name | description |
|---|---|
| suggestion | The raw Markdown fix suggestion generated by AI |
| confidence | Confidence score (0-100) |
| error-type | Detected error category (e.g., build-error, test-failure) |