Python checks you can trust¶
Catch complicated functions, type mistakes, and missing documentation before you merge a Python change.
pythonprs gives you a GitHub Actions workflow and the same checks on your own computer. It also tests deliberately bad examples, so a broken check cannot quietly look successful.
Start with your goal¶
-
Try the checks locally
Run a passing check, make a temporary example fail, and repair it.
-
See a GitHub run
Make a small change in your own fork and watch its pull-request checks.
-
Set up your project
Add the workflow and rules to an existing Python repository.
-
Protect your branch
Make passing checks a requirement before anyone merges a change.
-
Fix a failure
Find the cause of a failed job and run its check again.
-
Publish your documentation
Preview a Zensical site and publish it to your own GitHub Pages URL.
Learn, do, look up, understand¶
These pages follow Diátaxis: each kind of documentation serves a different need.
- Tutorials guide you through a complete practice exercise. Start locally, then try GitHub.
- How-to guides help you finish a task: install the workflow, require checks, repair failures, or publish documentation.
- Reference gives you exact check requirements, configuration settings, and commands.
- Explanation connects the ideas: how the checks work, what types and docstrings tell you, and real findings and repairs from iParq.
What a passing result means¶
Each function stays within two complexity limits. Type checks find mistakes that can be detected from the code's type information. Every function has its own documentation, with a summary above a minimum length. The tests of the checks themselves also pass.
A passing result does not prove that a program is correct or that its documentation is useful. You still review the change and test its behavior. GitHub prevents merging a failed check only after you configure required checks.
For the original tool research and review evidence, see the maintainer record.