Topolograph MCP

Topolograph MCP is an MCP server that connects AI assistants and LLMs to the Topolograph API for OSPF, IS-IS, and BGP network topology analysis. Network engineers, site reliability teams, and infrastructure operators use it to monitor routing states, calculate paths, and diagnose topological changes directly from chat interfaces. By bridging language models to network infrastructure data, the server allows agents to query nodes, edges, and subnets across multi-area networks. It also provides tools to inspect MPLS-TE LSP tunnels, run constrained shortest path (CSPF) calculations, and simulate edge failure impacts without modifying the live network. For incident response, it correlates network event streams into timeline waves to identify flap or outage patterns. The server supports read-only enforcement to guard against unauthorized mutations, while still offering BGP session auditing, route comparison, and VRF route resolution for Topolograph versions 2.69 and newer.

Category: Cloud & Infrastructure

Tags: is-is, networking, ospf, routing, topology

Visit Topolograph MCP

How to install and configure Topolograph MCP

Follow these steps to set up and run the Topolograph MCP server: 1. Install dependencies from the repository: bash pip install -r requirements.txt 2. Configure your environment variables: bash export TOPOLOGRAPH_API_BASE="https://your-topolograph-api-url" export TOPOLOGRAPH_API_TOKEN="your-api-token" export TOPOLOGRAPH_MCP_READ_ONLY="true" 3. Start the server directly: bash python mcp-server.py Alternatively, deploy via Docker Compose using the topolograph-docker stack: bash git clone https://github.com/Vadims06/topolograph-docker.git cd topolograph-docker docker-compose up -d 4. Connect your MCP client to the HTTP endpoint at http://0.0.0.0:8000/mcp.

What you can do with Topolograph MCP

• Calculate shortest paths and CSPF routes between network nodes while accounting for active MPLS-TE tunnels. - Group OSPF and IS-IS host or link up/down transitions into timeline waves to investigate outages. - Simulate single or multi-link edge failures to predict network traffic re-routing and blast radius. - Inspect BGP speaker RIB tables, compare route states between timestamps, and resolve VPN paths. - Retrieve node roles, router flags, and diagram edge health metrics during real-time network troubleshooting.

Key facts

• https://github.com/Vadims06/topolograph-mcp-server
• Cloud & Infrastructure, Monitoring & Observability
• is-is, networking, ospf, routing, topology

Part of MCP Servers

What can Topolograph MCP do?

Topolograph MCP lets AI models interact with the Topolograph API to inspect OSPF, IS-IS, and BGP networks. It calculates shortest paths, runs CSPF feasibility checks, simulates link failure impacts, retrieves node and edge topology data, and reviews network incident timelines.

How do I install Topolograph MCP?

You can install dependencies with pip install -r requirements.txt and execute python mcp-server.py after exporting the TOPOLOGRAPH_API_BASE environment variable. It can also be deployed alongside the Topolograph stack using docker-compose via the topolograph-docker repository.

Does Topolograph MCP support read-only mode?

Yes, Topolograph MCP defaults to read-only mode via the TOPOLOGRAPH_MCP_READ_ONLY environment variable. In this mode, mutation operations such as uploading graphs or modifying MPLS-TE LSP tunnels are hidden from advertised tools and blocked from execution.

Which MCP clients work with Topolograph MCP?

Topolograph MCP exposes an HTTP MCP endpoint on port 8000, making it compatible with any MCP client that supports remote HTTP transport, including custom agent frameworks, Claude Desktop, and IDE tools configured to connect to remote MCP endpoints.