Abaqus Control MCP
MCP serverCommunity
A local bridge into the open Abaqus/CAE session: runs CAE Python, follows solves, queries ODBs and returns viewport images. By Whfkl.
215 starsUpdated 5 weeks agoGitHub data from 5 Nov 2026
At a glance
- Type
- MCP server
- Works with
- Abaqus
- How it connects
- Drives a running copy of the tool
- Access
- Can modify models
- Tool versions
- Abaqus/CAE with Python 3 (a separate fork covers Python 2 releases such as 2021)
- Operating systems
- Windows, Linux
- Needs a Abaqus licence
- Yes
- Transport
- Local (stdio)
- AI clients documented
- Codex, Claude Code
- Licence
- MIT
- Language
- Python
What it can do
- Six tools - ping, run_python, set_workdir, monitor_job_status, inspect_odb and capture_viewport.
- Runs Python against the current mdb and session through a GUI plug-in on 127.0.0.1:48152, with an execution ID tying each result to its call.
- Combines CAE job status with .log, .sta, .dat and .msg evidence, and reports UNKNOWN when they conflict.
- Summarises a field by step, frame and set, or returns a sampled history curve, opening ODBs read-only.
Limitations
- run_python has the full power of the Abaqus scripting console; review generated modelling and job commands.
- You start the bridge from Plug-ins > Abaqus-Control-MCP > Start MCP Bridge each time CAE opens.
- A standalone server executable is offered only for Windows; other systems install from source.
Install
uv tool install git+https://github.com/Whfkl/Abaqus-Control-MCP.gitcodex mcp add abaqus-control-mcp --env ABAQUS_MCP_HOST=127.0.0.1 --env ABAQUS_MCP_PORT=48152 --env ABAQUS_MCP_TIMEOUT=120 -- abaqus-control-mcp-serverMCP servers and skills run code on your computer. Review the source before installing.
AI-generated models, drawings and results must be checked by a qualified engineer before use.