Skip to content

README.md

Latest commit

 

History

History
249 lines (179 loc) · 5.85 KB

README.md

File metadata and controls

249 lines (179 loc) · 5.85 KB
Raw

OpenFix 🤖

AI-powered GitHub issue solver that automatically discovers bugs, generates patches, and creates pull requests.

License: MIT Python 3.10+

✨ Features

  • 🔍 Automatic Issue Discovery - Scans repositories and triages issues by likelihood of automated fix
  • 🛠️ AI Patch Generation - Uses Gemini AI to generate code patches with validation
  • 🔄 Iterative Repair - Self-improving patches through validation feedback loops
  • 📊 Confidence Scoring - Risk assessment (0-100) for each generated patch
  • 🎯 Draft PR Creation - Automatically creates PRs with AI disclosure and review checklist
  • ⚡ GitHub Actions - Full CI/CD integration for automated workflows
  • 🖥️ Interactive CLI - Beautiful terminal UI for patch review and approval

🚀 Quick Start

Installation

git clone https://github.com/CodeRustyPro/openfix.git
cd openfix

pip install -r requirements.txt

export GEMINI_API_KEY="your-gemini-api-key"
export GITHUB_TOKEN="your-github-token"

Basic Usage

# 1. Discover issues in a repository
python agents/discovery/discover_issues.py https://github.com/owner/repo

# 2. Generate a patch for a specific issue (interactive)
python scripts/cli.py solve https://github.com/owner/repo --issue 123

# 3. Full automation (discover + patch + PR)
python scripts/automate_full_pipeline.py \\
  --repo-url https://github.com/owner/repo \\
  --create-pr

📖 Usage Guide

CLI Commands

Discover Issues:

python scripts/cli.py discover https://github.com/owner/repo --limit 10

Solve with Approval:

python scripts/cli.py solve https://github.com/owner/repo --issue 42
# Shows confidence score, diff preview, asks for approval

Auto-Approve (CI mode):

python scripts/cli.py solve https://github.com/owner/repo --issue 42 --no-confirm

Check Status:

python scripts/cli.py status

GitHub Action

Add .github/workflows/openfix.yml:

name: OpenFix
on:
  issues:
    types: [labeled]

jobs:
  fix:
    if: contains(github.event.issue.labels.*.name, 'openfix')
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - uses: CodeRustyPro/openfix-action@v1
        with:
          gemini-api-key: \${{ secrets.GEMINI_API_KEY }}
          github-token: \${{ secrets.GITHUB_TOKEN }}

Label any issue with openfix and the bot automatically generates a fix!

🔧 Configuration

Environment Variables

  • GEMINI_API_KEY - Required for AI patch generation
  • GITHUB_TOKEN - Required for PR creation, optional for read-only operations

Config File

Edit config/config.yml:

llm:
  model: "gemini-2.0-flash-exp"
  temperature: 0.3
  max_retries: 2

validation:
  enabled: true
  timeout: 300

patch:
  max_files: 5
  confidence_threshold: 60

📊 Confidence Scoring

OpenFix calculates a 0-100 confidence score for each patch:

Score Risk Recommendation
75-100 Low High confidence - Consider auto-merge
50-74 Medium Review recommended
0-49 High Manual review required

Scoring Factors:

  • Triage Priority (25%)
  • Issue Complexity (20%)
  • Validation Results (35%)
  • Repair Iterations (20%)

🧪 Testing

Sandbox Testing

# Create test repository
python scripts/setup_sandbox.py

# Follow instructions to create GitHub repo
# Then test OpenFix on it

Run Tests

# Unit tests
pytest tests/unit/

# Full test suite
pytest tests/ -v --cov=.

🛡️ Safety Features

  • ✅ Draft PRs Only - Never auto-merges
  • ✅ AI Disclosure - Clear attribution in every PR
  • ✅ Review Checklist - Ensures human oversight
  • ✅ Risk Labeling - High-risk patches clearly marked
  • ✅ Validation - Automated testing before PR creation

🏗️ Architecture

OpenFix Pipeline:
┌─────────────┐
│   Discover  │ → Scan repo for issues
└──────┬──────┘
       │
┌──────▼──────┐
│   Triage    │ → Filter & rank by suitability
└──────┬──────┘
       │
┌──────▼──────┐
│  Generate   │ → AI creates patch
└──────┬──────┘
       │
┌──────▼──────┐
│  Validate   │ → Run tests & checks
└──────┬──────┘
       │
┌──────▼──────┐
│   Repair    │ → Refine based on feedback
└──────┬──────┘
       │
┌──────▼──────┐
│ Create PR   │ → Draft PR with disclosure
└─────────────┘

🤝 Contributing

We welcome contributions! See CONTRIBUTING.md for guidelines.

  1. Fork the repository
  2. Create a feature branch (git checkout -b feature/amazing-feature)
  3. Commit your changes (git commit -m 'Add amazing feature')
  4. Push to the branch (git push origin feature/amazing-feature)
  5. Open a Pull Request

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

📧 Support

Made with ❤️ by CodeRustyPro