An intelligent DevOps agent that automatically diagnoses CI failures, proposes fixes, and validates solutions using AI-powered reasoning and tool execution.

The CodeOps Agent follows a sophisticated multi-step pipeline to diagnose and fix CI failures:

graph TD
    A[CI Failure Detected] --> B[Context Retrieval]
    B --> C[AI Planning & Analysis]
    C --> D[Tool Execution]
    D --> E[Code Patching]
    E --> F[Test Execution]
    F --> G{Success?}
    G -->|Yes| H[Verification & Documentation]
    G -->|No| I[Critic Agent Review]
    I --> J[Refined Planning]
    J --> D
    H --> K[Slack Notification]
    
    subgraph "Data Layer"
        L[FAISS Vector Store]
        M[SQLite Database]
        N[GitHub Repository]
    end
    
    subgraph "AI Components"
        O[OpenAI GPT-4]
        P[Sentence Transformers]
        Q[LangGraph Reasoning]
    end
    
    B --> L
    C --> O
    C --> Q
    D --> N
    H --> M

• Docker and Docker Compose
• Git
• OpenAI API Key
• GitHub Token (optional, for repository access)

1. Clone the repository:

   git clone <repository-url>
   cd codeops-agent

2. Configure environment:

   cp .env.example .env
   # Edit .env with your API keys

3. Start the application:

   docker compose up --build

4. Access the services:

   • Dashboard: http://localhost:5173
   • API: http://localhost:8000
   • Health Check: http://localhost:8000/healthz

1. Backend Setup:

   # Create virtual environment
   python -m venv .venv
   source .venv/bin/activate  # On Windows: .venv\Scripts\activate
   
   # Install dependencies
   pip install -r requirements.txt
   pip install sentence-transformers
   
   # Run backend
   uvicorn backend.app.main:app --reload --port 8000

2. Frontend Setup:

   cd frontend
   pnpm install
   pnpm run dev

The Docker setup includes:

• Backend (FastAPI): Port 8000, Python 3.11-slim
• Frontend (React): Port 5173, Node 22 with Vite
• FAISS Store: Vector database for semantic search
• Demo Repository: Mounted for testing purposes

The agent monitors CI pipelines and detects failures through webhooks or manual triggers.

• Analyzes failure logs and error messages
• Retrieves relevant code context using FAISS vector search
• Gathers repository metadata and recent changes

• Uses OpenAI GPT-4 to analyze the failure
• Generates a step-by-step fix plan
• Identifies required tools and resources

• Executes git operations (checkout, branch, commit)
• Runs tests to validate fixes
• Applies code patches and modifications

• Re-runs tests to confirm fixes
• Validates solution completeness
• Documents changes and rationale

• Sends results to Slack (if configured)
• Updates run logs and metrics
• Provides detailed trace information

# Trigger agent for run ID 1
curl -X POST http://localhost:8000/agent/run/1

# Get system metrics
curl http://localhost:8000/metrics

# List all runs
curl http://localhost:8000/runs

The dashboard provides real-time insights into:

• Success Rate: Percentage of successful fixes
• Average Latency: Time to complete fixes
• Confidence Trends: AI confidence scores over time
• Run Traces: Detailed execution logs

{
  "total_runs": 45,
  "successful_runs": 38,
  "success_rate": 0.844,
  "average_latency": 2.3,
  "confidence_trend": [0.85, 0.87, 0.89, 0.91]
}

# Backend tests
python -m pytest test_*.py

# Integration tests
python test_endpoints.py
python test_github_integration.py
python test_slack_integration.py

The included demo-repo contains a simple calculator with intentional bugs for testing:

# Run demo tests
cd demo-repo
python test_calculator.py

1. Docker Build Fails:

   • Ensure Docker is running
   • Check available disk space
   • Verify network connectivity

2. API Connection Issues:

   • Check if backend is running on port 8000
   • Verify CORS settings in backend/app/main.py
   • Check firewall settings

3. FAISS Index Issues:

   • Ensure faiss_index.bin and faiss_index_texts.pkl exist
   • Check file permissions
   • Rebuild index if corrupted

4. Environment Variables:

   • Verify .env file exists and is properly formatted
   • Check API key validity
   • Ensure no trailing spaces in values

# View Docker logs
docker compose logs -f backend
docker compose logs -f frontend

# View specific service logs
docker compose logs backend

# Build and tag images
docker build -f Dockerfile.backend -t codeops-agent/backend:latest .
docker build -f Dockerfile.frontend -t codeops-agent/frontend:latest .

# Push to registry
docker push codeops-agent/backend:latest
docker push codeops-agent/frontend:latest

• Use production-grade database (PostgreSQL)
• Configure proper secrets management
• Set up monitoring and alerting
• Enable HTTPS with proper certificates
• Configure backup strategies for FAISS index

• GitHub Integration
• Slack Integration
• LangGraph Integration
• Critic Agent Integration

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests
5. Submit a pull request

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

Ready to automate your DevOps workflow? Start with docker compose up and watch the magic happen! 🎉