Small-LLM, multi-stage bot that finds, patches & PRs bugs in your repository using github actions—just by adding one workflow file.
| Feature | Description |
|---|---|
| Drop-in | Copy a GitHub Actions workflow → the agent starts running in CI. |
| Tool-augmented | Fault localisation, patch synthesis, test-run validation inside one container. |
| Cost-aware | Uses smaller models by default with optional larger model fallback and cost controls. |
| Secure | All operations run within your CI environment with proper permission handling. |
-
Add a workflow to your repo at
.github/workflows/auto-fix.yml:- Copy the workflow file from this repository's
.github/workflows/auto-fix.yml - Copy the filter script from
.github/scripts/filter_issues.py
- Copy the workflow file from this repository's
-
Configure required secrets in your repository:
- Go to Settings > Secrets and variables > Actions
- Add
GOOGLE_API_KEYor other LLM provider API key
-
Enable PR creation permissions:
- Go to Settings > Actions > General
- Scroll to "Workflow permissions"
- Enable "Allow GitHub Actions to create and approve pull requests"
- Click "Save"
-
Customize behavior (optional):
- Create a
bugfix.ymlin your repository root - Override default settings (labels, max attempts, etc.)
- Create a
The BugFix Agent operates in sequential stages:
- Prepare: Sets up the environment and creates a new branch
- Localize: Identifies the source of the bug in your codebase
- Fix: Generates patches using LLM-based code synthesis
- Validate: Builds and tests the proposed fixes
- Apply: Commits changes to the branch
- Report: Creates a pull request with the fix
Each issue with the configured label will be processed automatically, with multiple fix attempts if needed.
- TODO complete this section
- .env for envrioment variables
- docker-compose for local development