Agent Skills
› Cai-aa/CAE-Agent-Hub
› socket-bridge
socket-bridge
GitHub通过Socket桥接连接Abaqus CAE，在Python内核中远程执行代码以进行固化模拟。涵盖连接检查、作业提交、状态监控及ODB结果提取等完整工作流，并详细说明了Python 2.7环境的特定语法限制。
触发场景
需要连接Abaqus CAE
在Abaqus内核中执行Python脚本
提交或监控仿真作业
读取Abaqus ODB结果文件
安装
npx skills add Cai-aa/CAE-Agent-Hub --skill socket-bridge -g -y
SKILL.md
Frontmatter
{
"name": "socket-bridge",
"description": "Connect to Abaqus CAE via socket bridge for curing simulation. Invoke when user needs to use socket bridge, connect to Abaqus, or execute Python in Abaqus kernel."
}
Socket Bridge Connection
Description
Connect to Abaqus CAE via socket bridge for curing simulation. Invoke when user needs to use socket bridge, connect to Abaqus, or execute Python in Abaqus kernel.
Content
Socket Bridge Overview
The socket bridge provides a TCP connection to Abaqus CAE at 127.0.0.1:48152, allowing remote execution of Python code in the Abaqus kernel.
Prerequisites
- Abaqus CAE must be open and running
- The socket bridge plugin must be active in Abaqus
Connection Check
Use the check_abaqus_connection MCP tool to verify the bridge status before executing any commands.
Available MCP Tools
| Tool | Description |
|---|---|
check_abaqus_connection |
Verify bridge connection status |
run_python |
Execute Python code in Abaqus kernel |
submit_job |
Submit a job and wait for completion |
monitor_job_status |
Check .sta/.msg file diagnostics |
inspect_odb |
Read ODB results and field outputs |
set_workdir |
Set the working directory |
list_jobs |
List existing jobs in the session |
run_python Usage
When using run_python, set the result variable to return data back to the caller:
# Example: get list of steps
result = []
for step_name in odb.steps.keys():
result.append(step_name)
Abaqus Python Environment Quirks
The Abaqus Python environment has several important quirks to be aware of:
Python Version
- Abaqus 2020+ uses Python 2.7 based environment
- Do not use Python 3 specific syntax
Unicode Issues
print()statements with unicode may cause errors- Use
result = ...to return data instead of printing
Imports
- Must import constants:
from abaqusConstants import * - This provides access to all Abaqus constants (OFF, ON, RUNNING, etc.)
String Formatting
- Use the
%operator for string formatting - Do NOT use
.format()or f-strings (not supported in Python 2.7)
# Correct
text = '%d nodes found' % node_count
# Incorrect (will fail)
text = '{} nodes found'.format(node_count)
text = f'{node_count} nodes found'
Dictionary Keys
- Plain strings work reliably as dictionary keys
- Unicode keys may fail in some operations
Standard Workflow
- Check connection: Use
check_abaqus_connectionto verify bridge is active - Set working directory: Use
set_workdirto configure the work directory - Create jobs: Use
run_pythonto create jobs from input files - Submit jobs: Use
submit_joborrun_pythonwithsubmit() - Monitor: Use
monitor_job_statusto check for errors or convergence issues - Copy ODB: Copy the ODB file from scratch to work directory
- Extract data: Use
inspect_odborrun_pythonto read results
版本历史
- cb4d869 当前 2026-10-19 04:34
