Get the FREE Ultimate OpenClaw Setup Guide →

openmm

OpenMM, an open-source platform for molecular dynamics (MD) simulations, is supported by an MCP server that offers a structured communication interface for task submission, management, and execution.

Installation
Run this command in your terminal to add the MCP server to Claude Code.
Run in terminal:
Command
claude mcp add --transport stdio phelanshao-openmm-mcp-server python run_openmm_server.py

How to use

The OpenMM MCP Server provides a natural language interface and programmable tools to configure, run, and monitor molecular dynamics simulations using OpenMM with optional Abacus DFT calculations. It exposes a set of basic and advanced tools to create MD simulations, perform DFT calculations, manage and monitor tasks, and analyze results. Typical workflows include setting up protein or membrane systems, running production MD simulations with various integrators, applying barostats and constraints, and performing enhanced sampling or metadynamics tasks. Use the available tools to craft step-by-step simulation setups and then issue commands to start, pause, or stop simulations, retrieve status, and analyze outcomes. The server is designed to run with GPU-accelerated OpenMM where available, and can fall back to CPU or reference implementations as needed.

How to install

Prerequisites:

  • Python 3.8+ and pip
  • Optional: Conda for OpenMM installation
  • Git

Installation steps:

  1. Clone the repository git clone <repository_url> cd openmm-mcp-server

  2. Create and activate a Python environment (recommended) python -m venv venv source venv/bin/activate # on Unix/macOS .\venv\Scripts\activate # on Windows

  3. Install Python dependencies pip install -r requirements.txt

  4. Install OpenMM (optional but recommended for real simulations) conda install -c conda-forge openmm

  5. Run a quick test to ensure the server starts python test_mcp_server.py

  6. Start the MCP server python run_openmm_server.py

Notes:

  • The repository uses a Python-based MCP server. If you plan to expose the server to external clients, consider configuring proper networking, authentication, and logging as described in the docs.
  • If you encounter a missing Python path error, ensure Python is in your PATH or use an absolute path to the Python executable.

Additional notes

Tips and common issues:

  • Environment variables help control behavior at runtime. Consider setting TASK_DATA_DIR for task persistence and MAX_CONCURRENT_TASKS to limit parallel tasks.
  • If you see an 'OpenMM not found' warning during startup, the server may run in mock mode; install OpenMM to enable real simulations.
  • Ensure the project directory has appropriate read/write permissions for the task_data directory and log files.
  • Use the test_mcp_server.py script to verify basic MCP functionality before running long simulations.
  • The OPENMM parameters reference and configuration options are documented in the repository's docs to tailor integrators, force fields, water models, platforms, nonbonded methods, constraints, barostats, and precision for your use case.

Related MCP Servers

Sponsor this space

Reach thousands of developers