一键导入
shell-script
Safely execute shell commands and scripts. Use when user asks to run a command, execute a script, or perform system operations via terminal.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Safely execute shell commands and scripts. Use when user asks to run a command, execute a script, or perform system operations via terminal.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
本地离线 OCR 文字识别,基于 tesseract.js。当用户发图片/截图/扫描件要求提取文字时使用,支持中英文。纯本地运行,无需联网。
Use when completing tasks, implementing major features, or before merging to verify work meets requirements
Use when implementing any feature or bugfix, before writing implementation code
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.
Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.
| name | shell-script |
| description | Safely execute shell commands and scripts. Use when user asks to run a command, execute a script, or perform system operations via terminal. |
| version | 1.0.0 |
| author | deskwand |
| license | MIT |
| metadata | {"tags":["Shell","Terminal","CLI","Scripts","System"]} |
| allowed-tools | shell,read |
Execute shell commands safely with risk classification.
read/write (bulk, recursive)Every shell call requires a risk level:
| Level | When | Examples |
|---|---|---|
low | Read-only, no side effects | ls, cat, git status, git diff, echo |
medium | Modifies state but reversible | mkdir, npm install, git add, git commit |
high | Destructive or irreversible | rm -rf, git push --force, chmod, system config changes |
high-risk commands are blocked by default. Use alternative approaches or get explicit user confirmation.
shell(
command="ls -la /path",
risk_level="low",
risk_reason="Read-only directory listing"
)
# Read-only inspection
git status --porcelain
ls -la
cat file.txt
wc -l file.txt
# Version checks
node --version
npm --version
# Build/test (no side effects outside workspace)
npm test
npm run build
# State changes — always confirm with user first
git commit -m "message" # confirm message
npm install <package> # confirm package name
rm file.txt # confirm file path
rm -rf without explicit confirmationgit push --force to shared brancheschmod 777 or permission escalationcurl output directly to shellWhen user provides a script file:
read the file first — understand what it doescd /path/to/project \
&& npm run lint \
&& npm test
Chain with && so failures stop execution. Never use ; for dependent commands.
sh or zshbrew vs apt-get, open vs xdg-open--no-verify, --force, or similar flags