Abaqus MCP Server
Enables Claude Code to interact with Abaqus FEM software for model generation, job submission, log monitoring, and result extraction from ODB files.
README
Abaqus MCP Server
A local stdio MCP server that enables Claude Code to interact with Abaqus FEM software โ model generation, job submission, log monitoring, and result extraction from ODB files.
๐ Ideal for simulation automation, parametric studies, post-processing pipelines, and teaching FEM scripting through natural language.
โจ Features
- 11 MCP tools covering the full Abaqus workflow โ from script generation to result extraction
- Workspace sandboxing โ all file operations are scoped to a configurable directory
- ODB extraction via generated Abaqus Python scripts (no direct binary parsing โ uses
odbAccess) - Security-first โ no arbitrary shell execution, path traversal blocked, confirmation required for destructive operations
- Windows support โ correctly handles Abaqus
.batfile execution viacmd.exe /c call - Output truncation โ prevents token blowout in LLM context windows
- 85 tests across unit, mocked integration, and MCP protocol layers
๐ Requirements
- Python 3.10+
- Abaqus (tested with Abaqus 2025; the
abaquscommand must be available) - The
mcpPython SDK (โฅ1.28.0) is installed automatically as a dependency
๐ Quick Start
# Clone the repository
git clone https://github.com/YOUR_USERNAME/abaqus-mcp-server.git
cd abaqus-mcp-server
# Install in editable mode with dev dependencies
pip install -e ".[dev]"
# Copy and edit environment config (optional)
copy .env.example .env
# Run the server
abaqus-mcp serve
๐ Claude Desktop Integration
Add to your claude_desktop_config.json:
{
"mcpServers": {
"abaqus": {
"type": "stdio",
"command": "python",
"args": ["-m", "abaqus_mcp_server.cli", "serve"],
"env": {
"ABAQUS_COMMAND": "abaqus",
"ABAQUS_MCP_WORKSPACE_DIR": "./abaqus_work"
},
"timeout": 600000
}
}
}
Or with Claude Code, add to your project's .mcp.json:
{
"mcpServers": {
"abaqus": {
"type": "stdio",
"command": "python",
"args": ["-m", "abaqus_mcp_server.cli", "serve"],
"env": {
"ABAQUS_COMMAND": "D:\\SIMULIA\\Commands\\abaqus.bat",
"ABAQUS_MCP_WORKSPACE_DIR": "${CLAUDE_PROJECT_DIR:-.}/abaqus_work"
}
}
}
}
โ๏ธ Configuration
All settings via environment variables (prefix ABAQUS_MCP_):
| Variable | Default | Description |
|---|---|---|
ABAQUS_COMMAND |
auto-detect | Path to Abaqus launcher (abaqus.bat) |
ABAQUS_MCP_WORKSPACE_DIR |
./abaqus_work |
Workspace root directory |
ABAQUS_MCP_MAX_CPUS |
4 |
Maximum CPUs per job |
ABAQUS_MCP_SCRIPT_TIMEOUT |
300 |
Script execution timeout (seconds) |
ABAQUS_MCP_JOB_TIMEOUT |
3600 |
Job execution timeout (seconds) |
ABAQUS_MCP_MAX_OUTPUT_CHARS |
20000 |
Max chars in a single tool output |
ABAQUS_MCP_ALLOW_OVERWRITE |
false |
Allow overwriting existing files |
ABAQUS_MCP_LOG_LEVEL |
WARNING |
Server-side logging level |
๐ ๏ธ Tools
| Tool | Description | Confirm? |
|---|---|---|
check_environment |
Verify Abaqus installation, Python paths, and workspace status | |
validate_workspace |
Create or validate workspace directory structure | |
generate_script |
Generate Abaqus Python modeling scripts (cantilever, static, modal, contact) | |
run_script |
Execute Python scripts via abaqus python with timeout control |
|
submit_job |
Submit an .inp file for FEM analysis |
โ |
job_status |
Check job progress from .sta and .lck files |
|
read_job_logs |
Read .sta/.msg/.dat/.log with automatic error/warning extraction |
|
extract_odb_summary |
Extract ODB metadata โ steps, frames, instances, field outputs | |
extract_field_output |
Extract field output (S, U, E, PEEQ, RF, etc.) as CSV with statistics | |
extract_history_output |
Extract history output (ALLIE, ALLSE, RF, etc.) as time-value CSV | |
clean_job |
Remove intermediate files, preserving .odb/.inp/.cae |
โ |
๐๏ธ Architecture
src/abaqus_mcp_server/
โโโ app.py # FastMCP singleton โ all tools register here
โโโ cli.py # argparse entry point
โโโ config.py # pydantic Settings (ABAQUS_MCP_ prefix)
โโโ security.py # Workspace sandboxing & path validation
โโโ abaqus_cli.py # Abaqus subprocess abstraction (all calls go here)
โโโ log_parser.py # .sta/.msg/.dat parsers
โโโ odb_script_templates.py # Generate Abaqus Python scripts for ODB extraction
โโโ constants.py # Status strings, file extensions, error keywords
โโโ tools/ # 11 MCP tool implementations (one per file)
Key Design Decisions
- FastMCP โ high-level MCP SDK with auto-generated JSON schemas from type annotations
- ODB extraction via generated scripts โ the server generates temporary Python scripts that run under
abaqus python(which hasodbAccess); system Python cannot importodbAccess - Windows subprocess โ Abaqus
.batfiles requirecmd.exe /c "call abaqus.bat ..."(noshell=True) - Workspace sandboxing โ
security.validate_path()enforces all file ops stay within the workspace - Confirmation gates โ destructive tools require
confirmed=Trueas an explicit user gate
๐ Security
- All file paths are validated against the workspace directory
- Path traversal (
../../) is blocked - Dangerous shell characters (
;,|,&,`,$, etc.) are rejected submit_job,run_script, andclean_jobrequire explicitconfirmed=True- No arbitrary shell execution โ all Abaqus calls use explicit argument lists
- Output is truncated to
ABAQUS_MCP_MAX_OUTPUT_CHARSto prevent context overflow
๐งช Testing
# Run all tests
python -m pytest tests/ -v
# Run a specific area
python -m pytest tests/test_security.py -v
# Run with coverage
python -m pytest tests/ --cov=abaqus_mcp_server --cov-report=html
Test Layers
| Layer | What | Requires Abaqus |
|---|---|---|
| Unit | security, config, log parsers | No |
| Mocked integration | AbaqusCLI, ODB extraction | No |
| MCP protocol | Server subprocess, tool listing | No |
| Live integration | Real Abaqus calls | Yes |
๐ Example: Cantilever Beam
examples/simple_beam/ contains a complete cantilever beam analysis:
beam.inpโ linear elastic beam with C3D8 elements- Ask Claude Code: "Submit the beam.inp and extract the maximum Mises stress"
The server will: validate the workspace โ submit the job โ check status โ extract ODB summary โ extract field output.
๐ Troubleshooting
Abaqus not found
Set ABAQUS_COMMAND to the full path:
set ABAQUS_COMMAND=D:\SIMULIA\Commands\abaqus.bat
ODB extraction fails
- The
.odbfile may be locked (job still running) โ check withjob_statusfirst - Abaqus Python must be able to read the workspace directory
- The generated temp scripts run under
abaqus python, not system Python
Path traversal errors
All file paths must be relative to ABAQUS_MCP_WORKSPACE_DIR.
๐ค Contributing
Contributions are welcome! See CONTRIBUTING.md for development setup, project structure, and guidelines.
๐ License
MIT โ see LICENSE for details.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.