🐍🔍 Python Patrol
Static analysis for Python performance anti-patterns & common mistakes — with inline PR fix suggestions.
Python Patrol is a zero-dependency GitHub Action that catches bugs and performance issues other linters miss. When it runs on your pull requests, it posts inline review comments with one-click fix suggestions — not just warnings, but actual fixes you can apply with a single click.
What Makes It Different?
Ruff and pylint are great. But they run locally and give you a list of warnings. Python Patrol runs in your PR and does something no other linter does well:
- Inline fix suggestions — clickable “Apply suggestion” buttons right on the diff
- Performance evidence — real CPython 3.12 benchmark data explaining why each pattern matters
- Zero dependencies — pure Python
astmodule, runs in ~100ms
Example
When Python Patrol finds a list membership check like if x in [1, 2, 3], it posts:
🐍 Python Patrol — PP002
Membership test against a list literal is O(n) — use a set literal instead.
📊 Why it matters: List membership is O(n) vs O(1) for sets. On a 1000-element collection, set lookup is ~100x faster (CPython 3.12 timeit).
💡 Suggested fix:if x in {1, 2, 3}:
15 Rules Across 3 Categories
Performance (PP001–PP006): Mutable defaults, list membership checks, string concat in loops, unnecessary list comprehensions, repeated attribute lookups.
Correctness (PC001–PC005): Bare except, mutable class variables, late-binding closures, assert in production, == None.
Security (PS001–PS003): Hardcoded secrets, eval/exec, pickle on untrusted data.
Quick Start
name: Python Patrol
on: [pull_request]
jobs:
patrol:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- uses: bBlazewavE/python-patrol@v1
with:
github-token: ${{ secrets.GITHUB_TOKEN }}
Links
Built with pure Python. No dependencies. Just better code reviews.
