Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mcp-server-langgraph.mintlify.app/llms.txt

Use this file to discover all available pages before exploring further.

​
Developer Onboarding Guide

Welcome to the MCP Server LangGraph project! This guide will get you up and running in under 10 minutes.

​
Prerequisites

Before you start, ensure you have:
  • ✅ Python 3.11-3.13 installed (python --version)
  • ✅ Docker Desktop or Docker Engine running (docker --version)
  • ✅ Docker Compose v2 installed (docker compose version)
  • ✅ Git installed (git --version)
  • ✅ uv (Python package manager) installed (installation)
Optional but recommended:
  • Node.js (for Mintlify docs): node --version
  • Make (usually pre-installed on Linux/macOS)

​
🚀 Quick Start (5 Minutes)

# 1. Clone the repository
git clone https://github.com/vishnu2kmohan/mcp-server-langgraph.git
cd mcp-server-langgraph

# 2. Run complete developer setup
make dev-setup
This will:
  • Install all dependencies
  • Start Docker infrastructure (OpenFGA, Postgres, Keycloak, Prometheus, Grafana, etc.)
  • Initialize OpenFGA authorization
  • Set up Keycloak SSO
  • Display next steps
Expected duration: 3-5 minutes ⏱️

​
Option B: Manual Setup

# 1. Install dependencies (uv sync creates .venv automatically)
make install-dev  # Same as: uv sync

# 2. Start infrastructure
make setup-infra

# 3. Initialize services (wait for services to be ready first)
sleep 10
make setup-openfga
make setup-keycloak
No manual venv creation needed! uv sync (run by make install-dev) automatically creates .venv and installs all dependencies.

​
✅ Verify Installation

# Check system health
make health-check
You should see:
  • ✓ All Docker services running
  • ✓ All ports responding (8080, 5432, 8180, 16686, 9090, 3000, 6379)
  • ✓ Virtual environment active

​
🏃 Running the Application

​
Run Unit Tests

make test-unit
Expected: All 437 tests passing ✅

​
Run the MCP Server (StreamableHTTP)

make run-streamable
The server will start on http://localhost:8000

​
Alternative: Run stdio MCP Server

make run

​
📊 Access Monitoring Dashboards

make monitoring-dashboard
This opens Grafana at http://localhost:3000 with:
  • Username: admin
  • Password: admin
Available Dashboards:
  • LangGraph Agent Performance
  • Security & Authentication
  • LLM Performance
  • SLA Monitoring
  • SOC2 Compliance
  • And 4 more…

​
🔧 Essential Configuration

​
1. Environment Variables

Copy the example environment file: 
cp .env.example .env
Update these required values in .env: 
# OpenFGA (get these from setup-openfga output)
OPENFGA_STORE_ID=<your-store-id>
OPENFGA_MODEL_ID=<your-model-id>

# Keycloak (get this from setup-keycloak output)
KEYCLOAK_CLIENT_SECRET=<your-client-secret>

# LLM API Keys (optional for testing)
ANTHROPIC_API_KEY=sk-ant-...
OPENAI_API_KEY=sk-...
GOOGLE_API_KEY=...

​
2. Re-run Setup if Needed

If you need to reset everything: 
make reset
This will clean up and re-initialize all services.

​
📚 Project Structure

mcp-server-langgraph/
├── src/mcp_server_langgraph/   # Main package
│   ├── core/                    # Agent, config, compliance
│   ├── auth/                    # Authentication & authorization
│   ├── llm/                     # LLM factory (LiteLLM)
│   ├── mcp/                     # MCP servers (stdio, HTTP)
│   ├── api/                     # REST API endpoints
│   └── observability/           # OpenTelemetry setup
├── tests/                       # Test suite (437 tests)
├── docs/                        # Mintlify documentation (77 pages)
├── deployments/                 # K8s, Helm, Kustomize configs
├── monitoring/                  # Grafana dashboards, Prometheus rules
├── examples/                    # Usage examples
└── Makefile                     # Developer commands

​
🧪 Running Tests

​
All Tests

make test

​
By Category

make test-unit          # Fast unit tests
make test-integration   # Integration tests
make test-compliance    # GDPR, HIPAA, SOC2, SLA tests
make test-property      # Property-based tests (Hypothesis)
make test-contract      # Contract tests (MCP protocol)

​
With Coverage

make test-coverage
Opens HTML report at htmlcov/index.html

​
Watch Mode

make test-watch
Re-runs tests on file changes (great for TDD)

​
🎯 Common Development Tasks

​
Code Quality

make format         # Format with black + isort
make lint           # Run flake8 + mypy
make security-check # Run bandit security scan

​
Pre-commit Hooks

make pre-commit-setup
This installs git hooks that automatically:
  • Format code with black
  • Sort imports with isort
  • Run linting (flake8)
  • Check types (mypy)
  • Scan for security issues (bandit)

​
View Logs

make logs-follow    # All services
make logs-agent     # Agent only
make logs-grafana   # Grafana only

​
Database Operations

make db-shell      # Open PostgreSQL shell
make db-backup     # Create backup
make db-restore    # Restore from backup

​
📖 Documentation

​
Serve Docs Locally

make docs-serve
Opens Mintlify docs at http://localhost:3000

​
Key Documentation

  • Architecture: docs/architecture/ (21 ADRs)
  • Deployment: docs/deployment/
  • Testing: docs/advanced/testing.mdx
  • Monitoring: monitoring/MONITORING_QUICKSTART.md
  • AI Agent Help: .github/CLAUDE.md, .github/AGENTS.md

​
🐛 Troubleshooting

​
Services Not Starting

# Check Docker is running
docker ps

# Check logs for errors
docker compose logs

# Restart services
make clean && make setup-infra

​
Port Conflicts

If ports 3000, 8080, 9090, etc. are in use: 
# Find process using port
lsof -ti:3000

# Kill process (replace 3000 with your port)
kill -9 $(lsof -ti:3000)

​
Tests Failing

# Run tests with more verbose output
pytest -vv --tb=long

# Run specific test
pytest tests/test_agent.py::test_specific_function -v

# Debug a failing test
make test-debug

​
Environment Issues

# Recreate virtual environment
rm -rf .venv
uv sync  # Creates .venv and installs dependencies

# Or use make
make install-dev

​
🎓 Learning Resources

​
Getting Started

  1. Read: README.md - Project overview
  2. Review: docs/getting-started/quickstart.mdx
  3. Explore: Architecture Decision Records in docs/architecture/

​
Understanding the Code

  1. Start with: src/mcp_server_langgraph/core/agent.py - LangGraph agent
  2. Then: src/mcp_server_langgraph/mcp/server_streamable.py - MCP server
  3. Review: src/mcp_server_langgraph/auth/middleware.py - Auth middleware

​
AI Assistant Configuration

  • Claude Code: .github/CLAUDE.md
  • GitHub Copilot: .github/copilot-instructions.md
  • Cursor AI: .cursorrules
  • OpenAI Codex: .openai/codex-instructions.md

​
🤝 Contributing

​
Before Making Changes

  1. Create a feature branch: git checkout -b feature/your-feature-name
  2. Install pre-commit hooks: make pre-commit-setup
  3. Run tests: make test-unit

​
Making Changes

  1. Write code following existing patterns
  2. Add tests for new functionality
  3. Run make format before committing
  4. Ensure make test passes

​
Submitting Changes

  1. Commit with conventional commits: feat:, fix:, docs:, etc.
  2. Push to your branch
  3. Open a Pull Request
  4. Ensure CI passes
See CONTRIBUTING.md for detailed guidelines.

​
🚀 Releasing

​
Automated Version Bumping

When a new GitHub release is created, deployment versions are automatically updated across all configuration files. Process:
  1. Create a new release on GitHub with a tag (e.g., v2.5.0)
  2. GitHub Actions automatically:
    • Updates pyproject.toml
    • Updates docker-compose.yml
    • Updates Kubernetes deployment manifests
    • Updates Helm chart version and appVersion
    • Updates Kustomize image tags
    • Commits changes to main branch
    • Adds comment to release with deployment commands
Manual Version Bump (if needed): 
# Test with dry run first
DRY_RUN=1 bash scripts/deployment/bump-versions.sh 2.5.0

# Apply version bump
bash scripts/deployment/bump-versions.sh 2.5.0

# Commit changes
git commit -am "chore: bump version to 2.5.0"
git push origin main
Deployment After Release:
  • Docker Compose: docker compose pull && docker compose up -d
  • Kubernetes: kubectl set image deployment/langgraph-agent langgraph-agent=langgraph-agent:2.5.0
  • Helm: helm upgrade langgraph-agent deployments/helm/langgraph-agent --set image.tag=2.5.0
  • Kustomize: kubectl apply -k deployments/kustomize/overlays/production

​
📞 Getting Help

​
Resources

​
Quick Commands Reference

make help              # Show all available commands
make dev-setup         # Complete setup
make quick-start       # Quick start
make health-check      # System health
make test-unit         # Run tests
make monitoring-dashboard  # Open Grafana
make logs-follow       # View logs
make clean             # Clean up

​
✨ Next Steps

  1. ✅ Complete setup: make dev-setup
  2. ✅ Run tests: make test-unit
  3. ✅ Explore dashboards: make monitoring-dashboard
  4. ✅ Read architecture docs: docs/architecture/overview.mdx
  5. ✅ Try examples: python examples/openfga_usage.py
  6. ✅ Make your first contribution!
Welcome to the team! 🎉 For questions, reach out via GitHub Discussions or open an issue. Happy coding! 🚀