Is any form of mocking allowed?

Yes, Level 1 Unit tests allow mocking internal environment details like system time, random seeds, and environment variables to ensure deterministic results.

What tools does this skill use for Level 2 Integration testing?

It utilizes Docker and subprocess-controlled binaries to run real services like Postgres or Redis locally, ensuring tests interact with actual infrastructure.

How are external API credentials handled for E2E tests?

It includes a pattern for loading credentials from environment variables and automatically skipping tests if the required tokens are missing, preventing test failures in limited environments.

Why does this skill discourage mocking external services?

Mocking external services often leads to tests that pass even when the real service would fail. This skill promotes dependency injection and real-world integration to ensure maximum confidence in the code.

How do I run only specific levels of tests?

The skill provides a pytest configuration that uses markers. You can run unit tests only with 'pytest -m "not integration and not e2e"' or target specific tiers directly.

Python Testing Strategy

Name: Python Testing Strategy
Author: simonheimlicher

bysimonheimlicher

•

安全与测试

Implements a high-confidence, three-tier testing methodology for Python applications using real dependencies and dependency injection.

关于

The testing-python skill for Claude Code enforces a robust testing framework based on three distinct levels: Unit, Integration, and E2E. It prioritizes 'Maximum Confidence' by discouraging the mocking of external services in favor of dependency injection and real-world testing environments like Docker. This skill provides Python-specific implementation patterns, data factories, and pytest configurations to ensure code reliability across different execution speeds and infrastructure requirements, moving away from brittle implementation-detail testing toward observable behavior validation.

主要功能

Implements strict mocking prohibitions for external services to ensure tests reflect reality
Includes Docker-based integration testing harnesses for databases and external binaries
Enforces a tiered testing model categorized by speed: Unit (<100ms), Integration (<1s), and E2E (<10s)
Provides standardized Dependency Injection (DI) patterns and protocols for Python logic
1 GitHub stars
Automates credential management and conditional skipping logic for environment-dependent E2E tests

使用场景

Integrating Docker containers into the local test suite to validate real database interactions
Setting up a new Python project with a standardized pytest structure and tiered markers
Refactoring legacy code to replace brittle mocks with robust, testable dependency injection

关于

主要功能

Implements strict mocking prohibitions for external services to ensure tests reflect reality
Includes Docker-based integration testing harnesses for databases and external binaries
Enforces a tiered testing model categorized by speed: Unit (<100ms), Integration (<1s), and E2E (<10s)
Provides standardized Dependency Injection (DI) patterns and protocols for Python logic
1 GitHub stars
Automates credential management and conditional skipping logic for environment-dependent E2E tests

使用场景

Integrating Docker containers into the local test suite to validate real database interactions
Setting up a new Python project with a standardized pytest structure and tiered markers
Refactoring legacy code to replace brittle mocks with robust, testable dependency injection