返回 Skill 列表
extension
分类: 开发与工程无需 API Key

pyx

使用pyx CLI安全执行Python代码。使用场景:运行Python、执行脚本、测试代码。触发词:pyx、run python、execute code、test script。默认模式:MANIFEST_IO(优先文件操作并带有清单输出)。

person作者: jakexiaohubgithub

pyx Executor

Use pyx for safe Python execution. Default: MANIFEST_IO mode.

Read-First: Environment (Mandatory)

If you need any of the following, you MUST read references/environment.md first:

  • Which Python packages are installed for pyx
  • OS details
  • Shell type/path and shell syntax notes

Do NOT do trial-and-error checks (e.g. guessing imports, probing commands). Prefer the generated environment document as the source of truth.

Note: pyx info can be slow. Prefer reading references/environment.md first.

Current Environment

  • OS: Windows (AMD64)
  • Shell: powershell
  • pyx Python: 3.12.12
  • pyx version: 0.1.0

Depends On (Soft)

Load these skills alongside pyx:

  • manifest - MANIFEST_IO contract and workflow
  • learn - skill extraction workflow and summary reference

MANIFEST_IO (Default)

pyx assumes MANIFEST_IO by default:

  • Read inputs from JSON files
  • Write outputs to files + a manifest
  • Print a short stdout summary (paths + sizes)
  • Check sizes before reading outputs into context

See the manifest skill for the full spec.

Non-Strict Mode (Opt-out)

Use only when user explicitly says:

  • "no strict mode"
  • "simple mode"
pyx run --file "temp/task.py"

Golden Rule

NEVER paste code inline to shell. Always write to file first:

# ❌ WRONG
pyx run --code "import os; print(os.listdir())"

# ✅ CORRECT
pyx ensure-temp --dir "temp"
# Write code to: temp/list_files.py
pyx run --file "temp/list_files.py"

Quick Reference

| Task | Command | |------|---------| | Create temp dir | pyx ensure-temp --dir "temp" | | Run script | pyx run --file "script.py" | | Run with input | pyx run --file "script.py" --input-path "input.json" | | Run with timeout | pyx run --file "script.py" --timeout 30 | | Run in directory | pyx run --file "script.py" --cwd "/path" | | Install package | pyx add --package "requests" | | Check environment | pyx info |

References

pyx-specific references:

Common use-cases: