What makes this different from automated unit tests?

This skill creates scripts for manual/exploratory verification of specific user stories that might be harder to capture in standard unit tests, making them reproducible and persistent for future regression.

Where are the test scripts stored?

Scripts are saved directly in your project's tests/manual/ directory, ensuring they stay with the codebase and can be run by any developer at any time.

What is diff-based validation?

It compares the actual output of a test against a pre-validated 'expected' file to ensure the behavior is exactly as intended, which is more reliable than simple heuristic checks.

Does it support UI testing?

Yes, it can generate Puppeteer-based scripts for UI verification in addition to curl-based scripts for direct API endpoint testing.

Can it modify the status of my tasks or stories?

No, the manual tester focuses strictly on providing evidence and a verdict; it does not change task statuses or create new tasks, keeping the workflow clean.

Manual Acceptance Tester

Name: Manual Acceptance Tester
Author: levnikolaevich

bylevnikolaevich

•

147

•

安全与测试

Automates the creation and execution of persistent bash-based manual test suites to verify story acceptance criteria.

The Manual Acceptance Tester (ln-522) bridges the gap between manual verification and automated regression by generating executable, idempotent bash scripts within the project's tests/manual/ directory. It parses Story Acceptance Criteria (AC) to build targeted checks using tools like curl and Puppeteer, implementing a strict fail-fast policy and diff-based validation against expected reference files. By standardizing test structure and documentation, it ensures that manual checks are reproducible, documented in project readmes, and integrated into a master test runner for consistent quality gate reporting.

主要功能

01Automatic project documentation updates including READMEs and test manifests

02AC-driven test generation for API (curl) and UI (Puppeteer)

03Strict fail-fast validation and diffing against expected reference outputs

04147 GitHub stars

05Persistent test suite structure with shared configurations and master runners

06Structured result reporting with pass/fail matrices for quality gates

使用场景

01Verifying new feature stories against specific Acceptance Criteria before merging

02Standardizing how manual testing evidence is collected and reported within a team

03Building a library of manual regression scripts for complex API or UI workflows

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add levnikolaevich/claude-code-skills ln-522-manual-tester

For use in Claude.ai and ChatGPT

Download Skill