jalapenolabs/Self Repair
Automatically diagnose and repair CI failures using LLM agents
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| error | Error message to diagnose. If omitted, the action automatically fetches failed step logs from the current workflow run via the GitHub API. | no | "" |
| engine | LLM engine (claude or codex) | — | claude |
| model | Model identifier to pass to the engine (e.g. claude-sonnet-4-6). Falls back to engine default. | no | "" |
| repo | GitHub repository (owner/repo) | — | ${{ github.repository }} |
| claude-api-token | Anthropic API key (required when engine is claude) | no | — |
| openai-api-token | OpenAI API key (required when engine is codex) | no | — |
| github-token | GitHub token for creating issues and PRs | — | ${{ github.token }} |
| max-turns | Maximum conversation turns per engine invocation | — | 50 |
| verbose | Enable verbose logging of prompts and engine output | — | false |
| version | Version of self-repair to use (e.g. "0.2.0", "latest") | — | latest |
Outputs
no outputs