Skip to content

Releases: CodeRustyPro/openfix

OpenFix v0.1.0 - AI-Powered Issue Solver

28 Nov 04:17
@CodeRustyPro CodeRustyPro
v0.1.0
ba35e11

Choose a tag to compare

View all tags
OpenFix v0.1.0 - AI-Powered Issue Solver Latest
Latest

OpenFix v0.1.0 - Release Notes

Release Date: November 27, 2025
Status: Initial Public Release

🎯 Vision

OpenFix transforms GitHub issue resolution from manual labor to intelligent automation. By combining AI-powered code generation with robust validation and human oversight, we're building the future of collaborative software maintenance.

Goal: Enable maintainers to spend less time on routine bugs and more time on innovation.

🏆 Technical Achievements

Core Pipeline

✅ End-to-End Automation - From issue discovery to PR creation
✅ AI-Powered Patch Generation - Using Google Gemini 2.0 Flash
✅ Iterative Repair Loop - Self-improving patches via validation feedback
✅ Confidence Scoring - 0-100 risk assessment with weighted factors
✅ Draft PR Creation - Automatic GitHub integration with AI disclosure

Developer Experience

✅ Interactive CLI - Beautiful Rich UI with patch preview and approval
✅ GitHub Actions Integration - Full CI/CD automation
✅ Comprehensive Logging - Production-grade error handling
✅ Unit Test Coverage - 16 tests across core components
✅ Sandbox Testing - Safe environment for validation

Code Quality

✅ Type Hints - Full type annotations for IDE support
✅ Error Handling - Robust with specific exception types
✅ Code Formatting - Black & isort applied
✅ Documentation - Professional README and contributing guidelines

🏗️ Architecture

┌─────────────────────────────────────────────────────┐
│                  OpenFix Pipeline                    │
├─────────────────────────────────────────────────────┤
│                                                       │
│  1. Discovery → Scan repo for open issues           │
│  2. Triage → Filter & rank by AI suitability        │
│  3. Generate → Create patch via Gemini AI           │
│  4. Validate → Run tests & static analysis          │
│  5. Repair → Refine based on feedback (iterative)   │
│  6. Score → Calculate confidence (0-100)             │
│  7. PR → Create draft with AI disclosure            │
│                                                       │
└─────────────────────────────────────────────────────┘

Key Components:

  • Solver Agent - Orchestrates the entire pipeline
  • Chunk Selector - Semantic code retrieval (FAISS/fallback)
  • LLM Client - Gemini API with rate limiting & retry logic
  • Confidence Scorer - Multi-factor risk assessment
  • PR Creator - GitHub integration with safety features
  • Validation Harness - Automated test execution

🎨 CLI Capabilities

Commands

Discover Issues:

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

Solve with Interactive Approval:

python scripts/cli.py solve https://github.com/owner/repo --issue 42
# Shows: confidence score, risk rating, diff preview
# Asks: approve patch? create PR?

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

Full Automation

python scripts/automate_full_pipeline.py \
  --repo-url https://github.com/owner/repo \
  --create-pr

🤖 GitHub Action Automation

Add .github/workflows/openfix.yml to any repository:

Features:

  • Trigger on issue labels (openfix)
  • Manual dispatch with inputs
  • Scheduled discovery (nightly)
  • Automatic artifact upload
  • PR confidence comments

Example workflow:

  1. User labels issue with openfix
  2. Action runs automatically
  3. Patch generated & validated
  4. Draft PR created if confidence > threshold
  5. Maintainer reviews & merges

🧪 Sandbox Testing Support

Create test repository:

python scripts/setup_sandbox.py

Generates:

  • calculator.py - Division by zero bug
  • user_manager.py - Null check bug
  • test_calculator.py - Unit tests
  • README.md - Documentation

Perfect for:

  • Testing OpenFix locally
  • Demonstrating capabilities
  • Validating new features

⚠️ Known Limitations

  1. Language Support - Currently optimized for Python projects
  2. Validation - Requires runnable test suite in target repo
  3. Complexity - High-complexity issues may require multiple repair iterations
  4. API Limits - Subject to Gemini API rate limits (handled gracefully)
  5. Token Costs - Large codebases consume more API tokens

🗺️ Roadmap for v0.2.0

Planned Features

Multi-Language Support:

  • JavaScript/TypeScript
  • Go
  • Rust

Enhanced Validation:

  • Static analysis integration (pylint, mypy)
  • Security vulnerability scanning
  • Performance regression detection

Learning System:

  • Outcome tracking database
  • Few-shot learning from past successes
  • Customizable prompt templates

UI Dashboard:

  • Web interface for monitoring
  • Real-time pipeline visualization
  • Historical metrics

Integrations:

  • Slack notifications
  • Discord webhooks
  • Email alerts

📊 Statistics

  • Lines of Code: ~3,000
  • Test Coverage: 16 unit tests
  • Dependencies: 8 core packages
  • Development Time: Phase 1-2 complete
  • License: MIT

🙏 Acknowledgments

Built with:

🚀 Get Started

# Install
git clone https://github.com/yourusername/openfix.git
cd openfix
pip install -r requirements.txt

# Configure
export GEMINI_API_KEY="your-key"
export GITHUB_TOKEN="your-token"

# Run
python scripts/cli.py discover https://github.com/owner/repo

Full Documentation: README.md

Made with ❤️ by the OpenFix Team

Assets 2
Loading