Python Environment Management Skill

Core Principles

1. Reuse Existing Environments - Don't recreate, reuse existing virtual environments
2. Use Project-Type Decision - Auto-select based on lock files
3. Recommend by Popularity - uv > pip > conda > venv
4. Minimize Interruption - Only ask when necessary

Tool Popularity Ranking

Decision Flow

┌─────────────────────────────────────┐
│  Detect project dependency files     │
└─────────────────────────────────────┘
              ↓
    ┌─────────┴─────────┐
    ↓                   ↓
  Clear decision       Unclear
    ↓                   ↓
  Use directly     Detect existing env
                        ↓
                  ┌─────┴─────┐
                  ↓           ↓
              Has env        No env
                  ↓           ↓
              Reuse      Assess complexity
                            ↓
                  ┌─────────┴─────────┐
                  ↓                   ↓
              Simple task       Needs deps
                  ↓                   ↓
            System Python      Recommend uv/conda

1. Clear Decisions (Execute Directly, No Ask)

When these files are detected, use the corresponding tool directly:

2. Detect Existing Environments (Reuse First)

# Priority: uv venv > conda > venv

# 2.1 Detect uv virtual environment
ls -la .venv/ 2>/dev/null && uv pip list 2>/dev/null | head -3

# 2.2 Detect conda environment
conda info --envs 2>/dev/null | grep "*" || echo $CONDA_PREFIX

# 2.3 Detect standard venv
ls -la venv/ .venv/ env/ 2>/dev/null

# 2.4 If exists → Reuse (activate and run commands)

Reuse Example:

Detected existing .venv/ directory
→ Activate: source .venv/bin/activate
→ Run: uv pip install <package>

3. When Unclear (Assess Complexity)

4. When to Ask User (Only These Cases)

✅ Ask:

1. Empty project + first dependency install → Ask which tool
2. Both requirements.txt + pyproject.toml → Ask which to use
3. User explicitly wants different tool → e.g., "I want conda"