GMAT MCP servers and AI agent skills

How AI connects to GMAT

An MCP server or skill can only do what GMAT lets programs do. GMAT exposes 2 automation interfaces:

• GMAT script language and GmatConsole (Input files): A GMAT mission is a text script: resources such as spacecraft, propagators, thrusters and burns, then a mission sequence of commands such as Propagate, BeginFiniteBurn and EndFiniteBurn, and Vary and Achieve inside a targeting loop. GmatConsole runs a script without the GUI, so an agent can write one, run it and read parse, convergence or run errors back. The syntax has traps (ElapsedSecs, not ElapsedSeconds; no parentheses in conditionals) that one server keeps as a list of idioms for the agent. Documentation
• GMAT Python API (gmatpy) (Python): GMAT installs include a Python binding in bin/gmatpy. Before first use you generate an API startup file by running BuildApiStartupFile.py from the api folder, and your Python version must match the binding. The Guy2Es server loads GMAT this way inside a separate worker process, so a GMAT crash fails one run rather than the whole server. Documentation

What NASA offers

GMAT is developed by NASA and documented on its project wiki, where the current release is R2026a. The wiki and download pages describe the application, its plugins and the Python API but no AI assistant or MCP server as of this check. One community server describes its aim as turning GMAT into a tool an agent can use rather than talk about.

Checked 4 Nov 2026. Sources: GMAT wiki (NASA), ginomoretta-creator/gmat-mcp-server README

What you can do with GMAT and AI

• Write a two-burn rendezvous or Hohmann transfer script, run it headless and let the agent fix parse or convergence errors from the classified output.
• Retarget burns with GMAT's differential corrector after a run shows the two-body solution misses under J2 gravity.
• Model a lunar orbit with a lunar gravity field and compare its long-term evolution with a reference ephemeris.
• Start from a NASA sample script for a known mission pattern instead of writing one from scratch.
• Export a 3D trajectory, ground track and the generated GMAT script after a run.

Taken from the listings' own documentation. Check each listing for exactly what it supports.

Installing

The listings here document setup for Claude Code, Claude Desktop and Cursor. Install steps, copied from each project's own documentation, are on each listing's page. Most GMAT integrations run on your own computer next to GMAT, so you need GMAT installed and licensed on the same machine unless the listing says otherwise.