Require fresh Python checks before an agent finishes.

Hooks that track Python edits and successful ruff, mypy and pytest evidence after those edits.

MIT

Software use follows the repository’s terms. Installation, integration and model usage are separate considerations.

View source & setup guide ↗Discuss setup or integration ↗

Usage pattern

Python edit
Successful checks
Completion evidence

A guide to how the project is used. Read its requirements and limits before installation.

Your starting point

An agent changing Python files and reporting completion.

A completion message can cite missing, failed or stale checks.

The capability to add

See that the required mechanical checks ran after the latest edit.

Installed hooks record edits and successful checks, then hold completion when required evidence is missing.

What’s in the project

A tool you can inspect.

Capabilities

Python edit tracking
Fresh ruff, mypy and pytest evidence
A stop check tied to recorded verification

Requirements

A supported Claude Code or Codex hook setup
ruff, mypy and pytest configured for your project

Important limits

Hooks must be installed, enabled and supported by the host.
Passing mechanical checks does not establish complete correctness, security or customer outcomes.

A useful fit

Consider this when…

Python teams who want completion claims backed by recent tool results.

A different starting point

Choose another approach when…

Non-Python projects or environments that cannot run the required hooks and checks.

Inspect the implementation

Public source, with a specific scope.

The repository and the files below show the implementation available to inspect. They do not establish results for your workflow.

Reviewed source snapshot: 4e90e0858e41.

Compatibility and behavior must be checked in your environment. No customer-result or quality-improvement guarantee is implied.

Code and implementation help

Choose the help you need.

Software. MIT. Read the repository’s terms and installation guide.

Setup and integration. Bring your current tools, a workflow and your desired result. Scope, fees, delivery and support are confirmed before work starts.

Operating costs. Existing CLI subscriptions, model-provider usage and hosting may apply depending on the tool and your configuration.

Compare published service scopes and prices ↗

Your next step

Inspect it. Then decide what to add.

Use the source and requirements to judge fit. If you want help adapting it to your workflow, start with a scoped conversation.

Discuss Python Evidence Gate setup ↗

No setup fee is agreed or payment taken on this page.