经 AI Skill Hub 精选评估,Abaqus智能建模控制器 获评「推荐使用」。这款MCP工具在功能完整性、社区活跃度和易用性方面表现出色,AI 评分 7.8 分,适合有一定技术背景的用户使用。
MCP协议开源工具,使Claude、Cursor等AI客户端可直接驱动Abaqus有限元软件进行参数化建模、仿真任务自动化。适合工程仿真、CAE分析、自动化建模的专业用户和开发者。
Abaqus智能建模控制器 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
MCP协议开源工具,使Claude、Cursor等AI客户端可直接驱动Abaqus有限元软件进行参数化建模、仿真任务自动化。适合工程仿真、CAE分析、自动化建模的专业用户和开发者。
Abaqus智能建模控制器 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
# 方式一:通过 Claude Code CLI 一键安装
claude skill install https://github.com/Whfkl/Abaqus-Control-MCP
# 方式二:手动配置 claude_desktop_config.json
{
"mcpServers": {
"abaqus-------": {
"command": "npx",
"args": ["-y", "abaqus-control-mcp"]
}
}
}
# 配置文件位置
# macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
# Windows: %APPDATA%/Claude/claude_desktop_config.json
# 安装后在 Claude 对话中直接使用 # 示例: 用户: 请帮我用 Abaqus智能建模控制器 执行以下任务... Claude: [自动调用 Abaqus智能建模控制器 MCP 工具处理请求] # 查看可用工具列表 # 在 Claude 中输入:"列出所有可用的 MCP 工具"
// claude_desktop_config.json 配置示例
{
"mcpServers": {
"abaqus_______": {
"command": "npx",
"args": ["-y", "abaqus-control-mcp"],
"env": {
// "API_KEY": "your-api-key-here"
}
}
}
}
// 保存后重启 Claude Desktop 生效
中文 | English
Let AI drive Abaqus directly. Describe the model you want — geometry, materials, loads, steps — and let the AI work inside your live Abaqus/CAE session.
Abaqus Control MCP connects Claude Code, Codex, Antigravity, and any other MCP-compatible client to a running Abaqus/CAE instance. You describe the task, the AI translates it into Abaqus actions, and the model updates in real time.
Older Abaqus versions ship Python 2. If your Abaqus uses Python 2, use the Python 2 compatible fork instead.
AI agent one-click install
If your AI agent supports natural-language installation, just prompt it with:
install https://github.com/Whfkl/Abaqus-Control-MCP
1. Install the package
You can install the package directly via Git, or from a cloned repository in your local directory.
Option A: Remote Git Install (direct install)
uv (recommended):
uv tool install git+https://github.com/Whfkl/Abaqus-Control-MCP.git
Or pip:
pip install git+https://github.com/Whfkl/Abaqus-Control-MCP.git
---
Option B: Local Directory Install (run inside the cloned project root)
uv (recommended):
uv tool install .
Or pip: pip install .
* Development Mode (editable install) If you plan to modify the source code and have changes apply dynamically:
uv (recommended):
uv tool install --editable .
Or pip: pip install -e .
Dependencies are declared in pyproject.toml — no requirements.txt needed. Any of the methods above will install the four CLI commands (abaqus-control-mcp-server, abaqus-control-check, abaqus-control-doctor, abaqus-control-setup) to your global environment or active virtual environment.
2. Install the GUI plugin
Run the following command to install automatically:
abaqus-control-setup
Alternatively, you can manually copy abaqus_mcp_bridge/gui_plugin.py from the project source to your Abaqus plugins directory (typically ~/abaqus_plugins/).
[!NOTE] You can override the target folder for both automatic and manual installations by setting the ABAQUS_MCP_PLUGIN_DIR environment variable.
3. Start Abaqus/CAE, activate the plugin
Plug-ins → Abaqus-Control-MCP → Start MCP Bridge
4. Configure your MCP client
Depending on the AI agent you use, add the MCP server configuration to the corresponding config file.
Add the following entry to the mcpServers section in ~/.claude.json:
"abaqus-control-mcp": {
"command": "abaqus-control-mcp-server",
"env": {
"ABAQUS_MCP_HOST": "127.0.0.1",
"ABAQUS_MCP_PORT": "48152",
"ABAQUS_MCP_TIMEOUT": "120"
}
}
Note: Add to themcpServerssection in~/.claude.json(global) or underprojects.<path>.mcpServersfor a specific project. Claude Code starts the MCP server automatically when you open a session.
Add the following entry to the mcpServers section in ~/.gemini/config/mcp_config.json. Specifying an absolute path is recommended (replace <Username> with your system username):
"abaqus-control-mcp": {
"command": "C:/Users/<Username>/.local/bin/abaqus-control-mcp-server.exe",
"env": {
"ABAQUS_MCP_HOST": "127.0.0.1",
"ABAQUS_MCP_PORT": "48152",
"ABAQUS_MCP_TIMEOUT": "120"
}
}
5. Verify
abaqus-control-check
| Variable | Default | Description |
|---|---|---|
ABAQUS_MCP_HOST | 127.0.0.1 | TCP bridge host |
ABAQUS_MCP_PORT | 48152 | TCP bridge port |
ABAQUS_MCP_TIMEOUT | 120 | Execution timeout in seconds |
ABAQUS_MCP_PLUGIN_DIR | ~/abaqus_plugins | GUI plugin target directory |
from abaqus_mcp_bridge.client import AbaqusBridgeClient
client = AbaqusBridgeClient(timeout=60)
result = client.execute("from abaqus import mdb; result = list(mdb.models.keys())")
print(result['return_value']) # ['Model-1', ...]
| Issue | Solution |
|---|---|
WinError 10061 Connection refused | The bridge service is not running in Abaqus/CAE. Please **first start Abaqus/CAE**, and select **Plug-ins -> Abaqus-Control-MCP -> Start MCP Bridge** from the top menu, then try again. |
Permission denied / os error 5 during install or update | An old abaqus-control-mcp-server process is still running in the background (used and locked by Claude Code/Cursor). Close the MCP client (such as Claude Code or Cursor) or manually terminate the process in Task Manager to release the file lock, then try installing again. |
MCP client says Connection closed / Server crashes with ModuleNotFoundError: No module named 'pydantic_core.core_schema' | The dependency packages (e.g. pydantic_core) in the default Python environment (e.g. Python 3.13) selected by the uv tool runner are corrupted or missing DLLs on Windows. Force a reinstall using another installed Python version: uv tool install --force --python 3.14 . |
command not found: abaqus-control-mcp-server | Reinstall or run abaqus-control-doctor |
| No output from server | Normal for stdio MCP server — it doesn't print to stdout |
Module abaqusGui can only be used in Abaqus/CAE GUI | Use **Plug-ins** menu, not File → Run Script |
| Connection timed out | Start the Abaqus plugin **before** the MCP server |
| Model doesn't appear in GUI | Run abaqus-control-check — verify "thread": "MainThread" |
| Claude Code doesn't see MCP tools | 1. Run claude mcp list to check if abaqus is registered. 2. If not listed, add the MCP config as described in Step 4. 3. Restart Claude Code after any config change. |
创新性强的专业工具,弥补AI与工程仿真软件的集成空白。代码质量待观察,Stars较少需关注长期维护能力。
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
✅ MIT 协议 — 最宽松的开源协议之一,可自由商用、修改、分发,仅需保留版权声明。
AI Skill Hub 点评:Abaqus智能建模控制器 的核心功能完整,质量良好。对于Claude Desktop / Claude Code 用户来说,这是一个值得纳入个人工具库的选择。建议先在非生产环境试用,再逐步推广。
| 原始名称 | Abaqus-Control-MCP |
| 原始描述 | 开源MCP工具:让 AI 可以驱动 Abaqus建模。Connect Claude, Cursor, and other MCP clients directly to you。⭐28 · Python |
| Topics | 有限元分析CAE自动化AbaqusMCP协议工程仿真 |
| GitHub | https://github.com/Whfkl/Abaqus-Control-MCP |
| License | MIT |
| 语言 | Python |
收录时间:2026-05-22 · 更新时间:2026-05-22 · License:MIT · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端