quality-validation

Validation systematically scores research and implementation plans against predefined quality rubrics, checking completeness, accuracy, citations, and actionability. It blocks progression if scores fall below thresholds, preventing low-quality outputs from reaching the planning or coding phase.

Validation detects incomplete API documentation, inaccurate endpoint descriptions, missing citations, and non-actionable guidance. It specifically verifies that implementation plans strictly adhere to documented APIs, preventing hallucinated or incorrect API usage.

Yes. Validation enforces mandatory checkpoints that automatically score each phase—research, planning, or coding—against objective rubrics. When quality gates fail, the workflow halts and directs refinement before proceeding to the next stage.

Reports provide structured scores for completeness, accuracy, citation quality, and actionability; a pass/fail status; specific defects identified; and actionable recommendations for improvement.

Run validation after generating research summaries for new features, before planning or coding begins. It applies to API research, Philosophy research, and Pattern research, catching issues early when they're cheaper to fix.

Validation applies to API research, Philosophy research, and Pattern research ResearchPacks. Each uses predefined rubrics tailored to that research type's quality standards and completeness criteria.