How is the test compliance score calculated?

The score is a 0-10 value derived from a weighted penalty system where isolation and determinism issues carry high severity (0.5 reduction), and anti-patterns carry medium severity (0.3 reduction).

What test anti-patterns does this skill detect?

It identifies 'The Liar' (tests without assertions), 'The Giant' (tests over 100 lines), 'Slow Poke' (tests exceeding 5s), 'Conjoined Twins' (unmocked unit tests), 'Happy Path Only', and 'Framework Testers'.

How does it improve test determinism?

The skill scans for common causes of flakiness such as unawaited timeouts, shared mutable state between tests, and reliance on non-deterministic values like Math.random() or system time without seeding or mocking.

Can it detect real database connections in tests?

Yes, it audits connection strings and setup blocks to identify real database usage (e.g., PostgreSQL/MySQL) and recommends switching to in-memory alternatives like SQLite or mocking the DB layer.

Test Isolation & Anti-Patterns Auditor

Name: Test Isolation & Anti-Patterns Auditor
Author: levnikolaevich

bylevnikolaevich

•

보안 및 테스팅

Audits test suites for isolation violations, determinism issues, and architectural anti-patterns to ensure reliable CI/CD pipelines.

소개

This specialized skill acts as an automated quality gate within the Claude Code environment, specifically designed to identify fragile or poorly structured tests. It systematically evaluates test files for external dependencies—such as unmocked APIs, real databases, or file system access—while detecting non-deterministic 'flaky' behavior caused by race conditions or shared mutable state. By identifying specific anti-patterns like 'The Liar' or 'The Giant' and providing a weighted compliance score, it empowers developers to build robust, fast, and maintainable test suites suitable for production-grade delivery workflows.

주요 기능

Comprehensive isolation auditing for APIs, databases, file systems, and network calls.
Detection of 6 major test anti-patterns including 'The Liar', 'The Giant', and 'Happy Path Only'.
Actionable refactoring recommendations with estimated effort levels (S/M/L).
38 GitHub stars
Deterministic check for flaky tests, race conditions, and order-dependent assertions.
Automated health scoring (0-10) based on issue severity and impact.

사용 사례

Debugging flaky test suites by identifying hidden race conditions and time-dependent logic.
Pre-commit auditing to prevent unmocked external dependencies from reaching the CI environment.
Refactoring legacy test suites to improve execution speed and maintainability by splitting 'Giant' tests.

소개

주요 기능

Comprehensive isolation auditing for APIs, databases, file systems, and network calls.
Detection of 6 major test anti-patterns including 'The Liar', 'The Giant', and 'Happy Path Only'.
Actionable refactoring recommendations with estimated effort levels (S/M/L).
38 GitHub stars
Deterministic check for flaky tests, race conditions, and order-dependent assertions.
Automated health scoring (0-10) based on issue severity and impact.

사용 사례

Debugging flaky test suites by identifying hidden race conditions and time-dependent logic.
Pre-commit auditing to prevent unmocked external dependencies from reaching the CI environment.
Refactoring legacy test suites to improve execution speed and maintainability by splitting 'Giant' tests.