CST Runtime CLI skills
SkillCommunity
Agent skills wrapping a 113-command CST CLI with guard rails against known CST traps, for modelling, solving, far fields and optimisation. By bbl21.
38 starsUpdated 5 months agoGitHub data from 5 Nov 2026
At a glance
- Type
- Skill
- Works with
- CST Studio Suite
- Tool versions
- CST Studio Suite 2026
- Operating systems
- Windows
- Needs a CST Studio Suite licence
- Yes
- AI clients documented
- Claude Code, Cursor, OpenCode
- Licence
- MIT
- Language
- Python
What it can do
- Gives the agent 113 JSON-schema commands for geometry, ports, mesh, parameters, sessions, 1D/2D results, far-field export and optimisation studies.
- A guard layer stops known mistakes, such as simulating without rebuilding after a parameter change or treating complex S11 as dB.
- Keeps the modeller session and the read-only results session strictly separate.
- Writes an audit trail of each call and produces self-contained HTML reports.
Limitations
- The guard layer only covers known patterns; the author advises human review at key steps.
- Needs Python 3.13 or later and uv; the skill text is in Chinese.
- Despite the repository name, it is a command-line toolchain used through skills, not an MCP server.
Install
Extract the release archive to your tool's skills directory (%USERPROFILE%\.config\opencode\skills\ or equivalent). The extracted structure should include skills/cst-runtime-cli/ and skills/cst-runtime-optimization/.Source: github.com/bbl21/cst-runtime-cli/blob/main/README.en.md
git clone https://github.com/bbl21/cst-runtime-cli.git
cd cst-runtime-cli
python skills/cst-runtime-cli/scripts/bootstrap.py --skill-path skills/cst-runtime-cli/scripts
uv run python -m cst_runtime list-toolsSource: github.com/bbl21/cst-runtime-cli/blob/main/README.en.md
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.