omni-mcp/isaac-sim-mcp

Overview

An MCP server and extension that connects large language models and AI agents to NVIDIA Isaac Sim for robotics simulation. It enables agentic control of simulation scenes, robot behaviors, and data generation through the Model Context Protocol.

Best for

Best for
Robotics developers and researchers who want to script or control Isaac Sim through AI agents and language models.

Use cases

• Controlling simulation environments programmatically via natural language or agentic AI
• Generating synthetic training data for robot perception and manipulation tasks
• Testing and iterating on robot control policies in a physics-accurate simulator

How to use

Install

uv pip install "mcp[cli]"

Tested with

Cursor

Example client config

{\n    "mcpServers": {\n        "isaac-sim": {\n            "command": "uv run /home/ubuntu/Documents/isaac-sim-mcp/isaac_mcp/server.py"\n        }\n    }\n}

Notes

An MCP server and extension that connects large language models and AI agents to NVIDIA Isaac Sim for robotics simulation. It enables agentic control of simulation scenes, robot behaviors, and data generation through the Model Context Protocol.

173 stars on GitHub. Last updated 2025-04-25. Licensed MIT.

Use cases

• Controlling simulation environments programmatically via natural language or agentic AI
• Generating synthetic training data for robot perception and manipulation tasks
• Testing and iterating on robot control policies in a physics-accurate simulator

Pros

• Open source with Python SDK for easy customization
• Leverages industry-standard MCP for interoperability with AI agents
• Directly integrates with a leading robotics simulator

Cons

• Requires NVIDIA Isaac Sim (proprietary, GPU-intensive) to be installed separately
• Small community and limited documentation (173 stars)
• Scope is narrow: only relevant to robotics simulation workflows

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.