pip-prune 🧹 — Optimize Your Python requirements.txt Automatically

Demo: pip-prune scanning and optimizing a Python project

pip-prune is a blazing-fast, zero-dependency Python CLI tool that scans your project, detects which packages in requirements.txt are actually used, and rewrites the file to the smallest, version-pinned, lint-clean set.

• Remove unused dependencies
• Pin versions for reproducibility
• Catch dynamic imports with runtime tracing
• Beautiful, colorized terminal output
• CI/CD and GitHub Actions ready

---

🚀 Use Case: Python Dependency Hygiene & Security

pip-prune is perfect for:

• Python developers who want to keep their requirements.txt minimal and accurate
• Teams enforcing dependency hygiene in CI/CD
• Open source maintainers who want reproducible, clean installs
• Anyone who wants to avoid dependency bloat and security risks

Keywords: python dependency cleaner, requirements.txt optimizer, remove unused python packages, pin python dependencies, python dependency hygiene, python requirements audit, python security, python reproducibility

---

⚡ Installation (pip)

pip install pip-prune

---

🛠️ Command Usage & Examples

1. Scan for Used Imports

pip-prune scan --paths src

• Lists all detected imports in your codebase.

2. Rewrite requirements.txt (Dry Run)

pip-prune rewrite --paths src --dry-run

• Shows what would be removed/kept, but does not modify files.

3. Actually Rewrite requirements.txt

pip-prune rewrite --paths src --no-dry-run --yes --inline

• Removes unused packages, pins versions, and updates your file (with backup).

4. Check for Unused Dependencies (CI/CD)

pip-prune check --paths src

• Exits 1 if changes are needed, 0 if clean (great for GitHub Actions).

5. Runtime Tracing for Dynamic Imports

pip-prune scan --paths src python3 main.py

• Runs your command and catches imports loaded at runtime (e.g., plugins, lazy imports).

6. Ignore Rules

pip-prune rewrite --ignore numpy,rich --dry-run

• Always keep specified packages, even if unused.

7. Config File Support

Create .pipprunerc.toml:

[pip-prune]
paths = ["src"]
ignore = ["numpy"]
tracer = false

Then run:

pip-prune rewrite --config .pipprunerc.toml --dry-run

---

📦 Example Workflow

1. Install:

   pip install pip-prune
   

2. Scan your project:

   pip-prune scan --paths myproject
   

3. Optimize requirements.txt:

   pip-prune rewrite --paths myproject --no-dry-run --yes --inline
   

4. Add to CI:

   - name: Check requirements
     run: pip-prune check --paths myproject
   

---

🧪 Testing & Development

• Run all tests:

  pytest
  

• Type checking:

  mypy pip_prune/
  

• Linting:

  ruff check .
  black --check .
  

• Build for PyPI:

  python -m build
  

---

🌟 Features

• Lightning-fast static AST scan
• Optional runtime tracing for dynamic imports
• Import-to-package resolution with heuristics
• Pretty Rich terminal output (color, icons, tables)
• Ignore rules via CLI, config, or env
• Config file support (pyproject.toml, .pipprunerc.toml)
• CI/CD and GitHub Actions ready
• 100% typed, mypy/ruff/black/isort clean
• Cross-platform: Linux, macOS, Windows (Python 3.9+)

---

📈 Why pip-prune?

• Save time: No more manual dependency audits
• Reduce risk: Fewer attack surfaces, less bloat
• Reproducibility: Pin exact versions for every install
• DevX: Beautiful, actionable output

---

📄 License

MIT — see LICENSE

---

👤 Author & Contact

Author: Sherin Joseph Roy
Email: sherin.joseph2217@gmail.com
GitHub: Sherin-SEF-AI/pip-prune

---

pip-prune: The fastest way to keep your Python dependencies clean, safe, and minimal.