LLSM (Local LLM for SCIA Model)
MCP serverCommunity
Builds, pushes and analyses SCIA Engineer 24 models from plain language through 25 OpenAPI tools, with a docked chat panel. By Quincy987.
0 starsUpdated 10 days agoGitHub data from 5 Nov 2026
At a glance
- Type
- MCP server
- Works with
- SCIA Engineer
- How it connects
- Drives a running copy of the tool
- Access
- Can modify models
- Tool versions
- SCIA Engineer 24.0
- Operating systems
- Windows
- Needs a SCIA Engineer licence
- Yes
- Transport
- Local (stdio)
- AI clients documented
- Claude Code
- Licence
- Not stated in its documentation
- Language
- Python
What it can do
- Exposes 25 SCIA tools to Claude Code through an MCP server that the repository's .mcp.json loads automatically.
- Creates models, rectangular sections, materials, nodes, beams, supports, load groups, load cases and line loads, then pushes them to SCIA and runs the analysis.
- Reads beam deflections and internal forces back from the results.
- Also offers a chat panel docked beside SCIA and a headless CLI that work with the Anthropic API, OpenAI, Gemini or a local OpenAI-compatible model.
Limitations
- Rectangular sections only; no point loads, surface loads, slabs or load combinations yet.
- Assumes a single SCIA installation on the machine.
- The repository has no licence file.
Install
git clone https://github.com/Quincy987/LLSM.git
cd LLSM
python -m venv .venv
.venv\Scripts\pip install -e .Run claude from inside the LLSM folder; Claude Code picks up the llsm MCP server from the repository's .mcp.json.Source: github.com/Quincy987/LLSM#option-a--claude-code-recommended-no-api-key
MCP 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.