Skip to content
Tresca

Code_Aster MCP servers and AI agent skills

One community MCP server works with code_aster, and it is read-only: Aster View reads MED result files (.med, .rmed) directly, without code_aster installed, and returns statistics, histories, line profiles, rendered images and journal figures. Nothing listed here writes command files or runs a solve. It has been tested on Linux only.

  • No official EDF (open source) server yet.
  • 1 MCP server listed.
EDF (open source)1 serverChecked 4 Nov 2026

Integrations for Code_Aster

Community and commercial servers

MCP serverCommunity

Aster Viewby tin7

A read-only MCP server and viewer for code_aster MED results: statistics, histories, line profiles, renders and journal figures.

LocalLinuxRead-only0· updated 7 days ago

Listings are sorted by origin, then by recent activity. GitHub data is refreshed nightly. Nothing here is ranked.

How AI connects to Code_Aster

An MCP server or skill can only do what Code_Aster lets programs do. Code_Aster exposes 3 automation interfaces:

  • Python command files (.comm) (Python): A code_aster study is a command file in Python: a sequence of commands, each documented in the user manual under its own number (DEFI_MATERIAU, for example, is U4.43.01). Because the file is plain Python, an agent can write or edit it like any script, then hand it to the solver. The test-case base on the project site is the usual starting point for a new study. Documentation
  • run_aster and the code_aster Python package (Command line): Recent versions are run through the run_aster script, and the developer documentation describes the code_aster and run_aster Python packages, including helpers for temporary folders and copying data and result files. A future MCP server that runs solves would most likely wrap run_aster. Documentation
  • MED result files (Input files): code_aster writes results in the MED format. Aster View reads these files directly with the h5py library, so results can be inspected on machines without code_aster or salome_meca, and exported to VTU for ParaView builds that lack a MED reader.

What EDF (open source) offers

code_aster is EDF's open-source finite element solver for mechanics, thermal analysis and dynamics, used alongside the salome_meca platform and developed in the open on GitLab. The project site announces user meetings and releases but no AI assistant or MCP server as of this check, so the only integration listed here is a community viewer.

Checked 4 Nov 2026. Sources: code_aster home page, code_aster source on GitLab

What you can do with Code_Aster and AI

  • List the meshes, groups, fields and time steps in a .rmed file and flag a truncated result.
  • Report the maximum von Mises stress over a named group, with its location and percentiles, using Gauss-point values.
  • Extract a time history at a node or a stress profile along a line, with a journal-ready plot.
  • Render a cut plane or threshold view to PNG and convert results to VTU for ParaView.
  • Build a journal-quality figure of a field on the deformed shape from a saved recipe, so the same view can be regenerated after each new run.

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

Installing

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

Questions

Is there a code_aster MCP server?

Yes, Aster View, a community project. It reads MED results only and never runs code_aster or changes your files.

Can Claude run a code_aster simulation?

Not through a listed server. Claude can help write a .comm command file, but you run it yourself with run_aster or salome_meca, then use Aster View to read the results.

Do I need code_aster installed to use Aster View?

No. It reads MED files with h5py and renders with VTK, so you only need Python 3.10 or later.

How should I install Aster View on Linux?

The project offers a one-line install script for Ubuntu and Debian that creates a virtual environment, installs the package and links the aster-view and aster-mcp commands. Then register aster-mcp with Claude Code using its full path. A VS Code extension is optional.

AI-generated models, drawings and results must be checked by a qualified engineer before use.

Sources (checked 4 Nov 2026): code_aster home page · code_aster developer documentation · tin7/aster-view README · Code_Aster product page

Code_Aster is developed by its own project (EDF (open source)). Tresca is independent and not affiliated with or endorsed by the Code_Aster project.