OpenSees MCP servers and AI agent skills

How AI connects to OpenSees

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

• OpenSeesPy (Python): OpenSeesPy is the Python interpreter for OpenSees: the same model, element, material, analysis and recorder commands as the Tcl version, called as ops functions. Because it is a pip package rather than a desktop program, an MCP server can import it and run analyses in-process with no GUI or licence. On Apple Silicon the opensees-mcp README says you need an x86_64 Python under Rosetta 2. Documentation
• OpenSees Tcl interpreter (Tcl): OpenSees also runs Tcl scripts, and GiD's gidopensees problem type writes its models as Tcl. The OpenSees Studio skills define a command-by-command mapping from Tcl to OpenSeesPy so translated models can be checked numerically against the originals. Documentation

What Open source (UC Berkeley / PEER) offers

OpenSees is an open-source framework from UC Berkeley and PEER rather than a commercial product, and we found no AI assistant or MCP server announced by the project. The servers and skills listed here are independent community work, and none comes from the OpenSees developers.

Checked 4 Nov 2026. Sources: OpenSees (UC Berkeley), OpenSeesPy documentation

What you can do with OpenSees and AI

• Create a 2D or 3D OpenSeesPy model from a description (nodes, elements, materials, boundary conditions) and run it without opening any GUI.
• Look up the exact signature and arguments of an OpenSeesPy or OPSVIS command before the agent writes analysis code.
• Translate an OpenSees Tcl script or a GiD gidopensees model into OpenSeesPy and compare the outputs within a tolerance.
• Draft response spectrum or time-history procedures following ASCE 7-22 or Latin American seismic codes, with every coefficient cited or flagged for checking.

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

Installing

The listings here document setup for Any client that reads SKILL.md, Claude, Claude Desktop, Cursor, Gemini and Google Antigravity. Install steps, copied from each project's own documentation, are on each listing's page. Most OpenSees integrations run on your own computer next to OpenSees, so you need OpenSees installed and licensed on the same machine unless the listing says otherwise.