Name: Python Assistant
Author: jvy

據 AlphaMoat 統計,截至 2026-09,Python Assistant 累計下載 4,574 次、獲 3 個星標,在全部 5,078 個 AI 技能中排名第 2,090 位,在「開發程式設計」分類(577 個)中排名第 268 位。

累計安裝

547

累計下載

4574

星標

全站排名

#2,090

全站前 42% · 「開發程式設計」前 47%

平台	累計安裝	累計下載	星標	評論
ClawHub ↗	69	2390	-	-
SkillHub ↗	547	4574	3	—

各平台的計數口徑不同,統計時間不一致。

這個技能是做什麼的

覆蓋專案初始化、依賴安裝、指令碼執行與環境排障，主張用專案內的虛擬環境而不是全域性安裝，用模組方式呼叫 pip 以避免直譯器不匹配。安裝前先看依賴清單檔案，遇到不認識的安裝鉤子或隨機指令碼要先徵得同意再執行。目標是讓本地與開發環境可復現、風險低。若 --venv 路徑設錯，recreate 流程可能刪除任意可寫目錄。使用明確位於專案內的 .venv，避免絕對路徑或父目錄，並在刪除/重建前驗證目標。

資料釋出於 2026-10-27

可以用來做什麼

給新專案搭一套可復現的環境
排查依賴裝不上或版本衝突

什麼時候用

要初始化 Python 專案時
環境報錯想按穩妥步驟排查時

示例提示詞

以下是根據技能文件生成的示例,不代表實際執行結果。

幫我給這個專案建虛擬環境並裝依賴

這個依賴衝突該怎麼排查

月度資料趨勢

累計下載4574

星標3

2026-09 起統計方法調整:同名技能按作者分別統計,並覆蓋全部技能。圖中虛線為調整前的月份,與之後的資料不直接比較,也不計算跨越調整的增長。

月份	累計下載	環比	星標
2026-09	4574	—	3
2026-08	3813	+75.2%	3
2026-07	2176	—	-
2026-05	2785	—	2
2026-03	1577	—	-

SKILL.md 原文

以下是作者釋出的原始檔案,按純文字顯示,本站未做修改或稽核;其中的連結不可點選,命令請核實後再執行。抓取於 2026-10-23。

2.7 KB · ClawHub v1.0.0

---
name: python
description: Use Python for practical project setup, dependency install, script execution, and environment troubleshooting with safe defaults. Use when tasks involve pyproject.toml or requirements.txt, virtual environment setup, running Python scripts/tests, packaging basics, or fixing common Python errors (interpreter mismatch, pip resolver conflicts, missing modules, build failures).
metadata:
  openclaw:
    emoji: "🐍"
    requires:
      bins: ["python3"]
      anyBins: ["pip", "pip3"]
    install:

展開全部(共 91 行)

      - id: "brew"
        kind: "brew"
        formula: "python"
        bins: ["python3"]
        label: "Install Python + pip (brew)"
      - id: "apt"
        kind: "apt"
        package: "python3-pip"
        bins: ["pip3"]
        label: "Install pip (apt)"
---

# Python

Use this skill to keep Python workflows reproducible and low-risk across local/dev shells.

## Safety Defaults

- Prefer project-local virtual environments (`.venv`) over global installs.
- Prefer `python3 -m pip ...` to avoid interpreter and pip mismatch.
- Inspect dependency files before install (`requirements*.txt`, `pyproject.toml`).
- Avoid executing unknown setup hooks or random install scripts without user approval.

## Standard Workflow

1. Detect current environment:

```bash
python3 --version
python3 -c "import sys; print(sys.executable)"
{baseDir}/scripts/python_env_tool.py doctor
```

2. Create or refresh a venv:

```bash
{baseDir}/scripts/python_env_tool.py bootstrap --venv .venv --requirements requirements.txt
```

3. Install project package (if `pyproject.toml` exists):

```bash
{baseDir}/scripts/python_env_tool.py install --venv .venv --editable
```

4. Run tests/tools from the venv interpreter:

```bash
.venv/bin/python -m pytest -q
.venv/bin/python -m pip list --outdated
```

## Task Recipes

```bash
# Install specific packages into venv
{baseDir}/scripts/python_env_tool.py install --venv .venv --package requests --package pydantic

# Install from requirements file
{baseDir}/scripts/python_env_tool.py install --venv .venv --requirements requirements-dev.txt

# Recreate corrupted venv from scratch
{baseDir}/scripts/python_env_tool.py bootstrap --venv .venv --recreate --requirements requirements.txt
```

## Troubleshooting Rules

- `ModuleNotFoundError`: verify command is run via `.venv/bin/python`, then reinstall deps.
- `externally-managed-environment`: stop global install attempts; use venv.
- Build failures on native deps: upgrade `pip setuptools wheel`, then retry.
- Multiple Python versions: always print and confirm `sys.executable` before fixes.

## Bundled Helper

Use the helper for repeatable environment setup and diagnosis:

```bash
{baseDir}/scripts/python_env_tool.py --help
{baseDir}/scripts/python_env_tool.py doctor
```