Can this skill help with flaky tests?

Yes, it provides strategies for removing time-dependencies, improving async handling, and using deterministic test data to eliminate random test failures.

How does it improve AI agent verification?

It emphasizes 'Three-Source Verification,' which cross-references internal event stores, external APIs, and ground-truth system states to confirm agent success.

Does this work for languages other than Python?

While the examples often feature Python and Pytest, the architectural principles, verification levels, and CI/CD strategies are applicable to any testing ecosystem, including Jest, Vitest, and others.

What is the 5-level verification hierarchy?

It is a framework used to evaluate test quality, moving from simple 'existence' checks (Level 1) to 'invariant' properties (Level 5) to ensure AI agents truly perform the tasks they claim.

Testing Architecture Expert

Name: Testing Architecture Expert
Author: AgentParadise

byAgentParadise

0•

セキュリティとテスト

Analyzes repository testing structures, identifies verification gaps, and plans comprehensive improvements with a focus on AI agent reliability.

The Testing Expert skill equips Claude with the specialized knowledge required to audit, architect, and improve software testing suites. It moves beyond basic unit testing by introducing a unique 5-level verification hierarchy specifically designed for AI agents, ensuring that code doesn't just claim to work but actually modifies the system state correctly. From hunting down 'assert True' anti-patterns to implementing complex fixture factories and three-source verification, this skill allows Claude to function as a senior QA architect capable of transforming weak test suites into robust, high-confidence verification systems.

主な機能

01Advanced testing patterns including Property-Based Testing and Three-Source Verification

025-Level Verification Hierarchy for high-confidence AI agent and system testing

03Automated discovery and structural analysis of existing test suites and CI configurations

040 GitHub stars

05Detection of common anti-patterns such as 'trust without verify' and useless assertions

06Comprehensive assessment reporting for identifying coverage gaps and prioritizing improvements

ユースケース

01Designing robust verification frameworks for AI agents that interact with external APIs and file systems

02Auditing legacy repositories to identify missing critical path tests and weak verification logic

03Optimizing CI/CD pipelines by implementing test markers, parallelization, and coverage gates

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add agentparadise/agentic-primitives testing-expert

For use in Claude.ai and ChatGPT

主な機能

01Advanced testing patterns including Property-Based Testing and Three-Source Verification

025-Level Verification Hierarchy for high-confidence AI agent and system testing

03Automated discovery and structural analysis of existing test suites and CI configurations

040 GitHub stars

05Detection of common anti-patterns such as 'trust without verify' and useless assertions

06Comprehensive assessment reporting for identifying coverage gaps and prioritizing improvements

ユースケース

01Designing robust verification frameworks for AI agents that interact with external APIs and file systems

02Auditing legacy repositories to identify missing critical path tests and weak verification logic

03Optimizing CI/CD pipelines by implementing test markers, parallelization, and coverage gates