AgentSUMO

Official

by urban-ai-institute · Python

MCP server for LLM-driven SUMO traffic simulation, scenario design, and result analysis.

MCP server for LLM-driven SUMO traffic simulation, scenario design, and result analysis. It implements the Model Context Protocol to integrate with development tools in support of workflows involving SUMO, urban mobility, and decision-support use cases informed by large-language-models.

🛠️ Key Features

Model Context Protocol (MCP)
LLM-driven SUMO traffic simulation
Scenario design
Result analysis

🚀 Use Cases

SUMO traffic-simulation scenario creation
Urban-mobility decision support
Agentic workflows using large-language-models

⚡ Developer Benefits

Focused integration point for MCP-based model-context interactions
Fits open-source development involving agentic-framework and decision-support

⚠️ Limitations

Description does not provide additional details about supported tools, configuration, or runtime constraints.

Topics

agentic-frameworkdecision-supportlarge-language-modelsmodel-context-protocolopen-sourcesumotraffic-simulationurban-mobility

Install

Configuration

Environment variables

SUMO_HOMErequired

Absolute path to the local SUMO installation directory (must contain bin/, share/, tools/).

AGENTSUMO_MCP_OUTPUT_BASE

Base directory for simulation outputs (networks, trips, results). Defaults to the current working directory.

Add to VS Code Add to Cursor

claude_desktop_config.json

json

{
  "mcpServers": {
    "agentsumo-mcp": {
      "command": "uvx",
      "args": [
        "agentsumo-mcp"
      ],
      "env": {
        "SUMO_HOME": "<YOUR_SUMO_HOME>",
        "AGENTSUMO_MCP_OUTPUT_BASE": "<YOUR_AGENTSUMO_MCP_OUTPUT_BASE>"
      }
    }
  }
}