This MCP server provides intelligent access to Nautobot APIs via semantic search and dynamic API requests. It exposes tools to perform CRUD operations against Nautobot endpoints, discover endpoints semantically, and search through a knowledge base of Nautobot documentation and code. You can run the server in stdio mode for local development or in HTTP/REST mode for web-based integrations. The included tools include nautobot_dynamic_api_request for arbitrary HTTP methods, nautobot_openapi_api_request_schema to discover available endpoints, and nautobot_kb_semantic_search to query the indexed Nautobot documentation and repositories. The server supports multi-environment Nautobot instances (dev, staging, prod) and uses a ChromaDB-backed vector store to enable fast semantic search with Git-based knowledge base updates.

Prerequisites:

• Python 3.11+ (3.11 recommended)
• Git
• Access to a Nautobot instance (token-based authentication)

Option A — Local Python Installation:

1. Clone the repository: git clone <repository-url> cd nautobot_mcp
2. Create and activate a virtual environment (optional but recommended): python -m venv venv

   On Windows: venv\Scripts\activate

   On macOS/Linux: source venv/bin/activate

3. Install dependencies: pip install -e .
4. Configure environment variables (see .env.example): cp .env.example .env

   Edit .env as needed

5. Run the server (stdio mode): python main.py --mode stdio

Option B — Docker (Recommended):

1. Clone the repository and navigate to nautobot_mcp.
2. Copy and customize environment variables: cp .env.example .env
3. Run in stdio mode with Docker Compose: docker compose up -d

   For HTTP mode, set MCP_TRANSPORT=http and MCP_PORT as needed, then rerun with the environment updated

4. View logs: docker compose logs -f

Tips and common issues:

• Ensure NAUTOBOT_TOKEN and NAUTOBOT_ENV reflect your Nautobot instance and environment.
• If using the knowledge base indexing features, ensure GITHUB_TOKEN is set if you plan to access private repositories used for indexing.
• For HTTP mode, open port 8000 (or as configured) and set MCP_TRANSPORT=http in the environment.
• The default configuration assumes local development; adapt SERVER_NAME, API_PREFIX, and repository settings as needed for production deployments.
• If you encounter SSL issues in development, you can set SSL_VERIFY=False in the environment, but enable SSL_VERIFY=True in production.
• The tools rely on a ChromaDB store; ensure the volumes or paths for chroma_db and models are writable and persisted between runs.