ag-criar-cenario-qat

You can automate QAT scenario creation by providing a feature and persona, which generates a four-layer specification (L1-L4) with a golden sample and anti-patterns aligned with existing rubrics and templates.

Anti-patterns in QAT specifications are generated negative examples that define incorrect implementation behavior. They are produced alongside a golden sample within the four-layer design to guide and validate quality assessments.

Generating QAT scenarios requires a specific project structure including a tests/qat/ directory, a qat.config.ts file, and rubrics under tests/qat/rubrics/v2/ to properly align and validate the generated scenarios.

The four-layer spec design structures QAT scenarios into levels L1 through L4, integrating acceptance criteria with existing rubrics. It outputs a golden sample and anti-patterns to comprehensively cover design and validation.

You can use existing rubrics to validate generated QAT scenarios. The generation process targets rubrics located in tests/qat/rubrics/v2/ to ensure the output scenarios align with established quality assessment criteria.

Structuring user stories for QAT generation requires including a clear feature description and a persona. This inputs allows the system to accurately translate requirements into four-layer specifications with golden samples and anti-patterns.