Agentic AI Workflow Engine

A powerful workflow engine for orchestrating AI agents using LangGraph, Pydantic Agents, and React Flow.

Features

• Visual workflow designer using React Flow
• LangGraph-based workflow orchestration
• Pydantic Agents with LLM capabilities
• MCP Server for tool execution
• Human-in-the-loop tasks
• Role-Based Access Control (RBAC)
• Workflow versioning and rollback
• PostgreSQL state management

Project Structure

langgraph-workflows/
├── backend/         # FastAPI backend service
├── mcp_server/     # Tool execution server
├── frontend/       # React Flow UI
├── docker/         # Docker configurations
├── k8s/            # Kubernetes manifests
└── docs/           # Documentation

Getting Started

Prerequisites

• Node.js >= 18
• Python >= 3.9
• PostgreSQL >= 14
• Docker (optional)
• Kubernetes (optional)

Development Setup

1. Frontend Development

cd frontend
npm install
npm run dev

2. Backend Development

cd backend
python -m venv venv
source venv/bin/activate  # or `venv\Scripts\activate` on Windows
pip install -r requirements.txt
uvicorn app.main:app --reload

3. MCP Server Development

cd mcp_server
python -m venv venv
source venv/bin/activate  # or `venv\Scripts\activate` on Windows
pip install -r requirements.txt
uvicorn app.main:app --reload --port 8001

4. Run the tests

uv run pytest tests/integration_tests/test_graph.py::test_self_learning_graphiti_agent_simple_passthrough -v

License

MIT License