Purpose: Datasets, executable cases, scoring rubrics, and run reports backing docs/06-ai-agents-rag/ai-evaluation-strategy.md and docs/09-quality-evaluation/ai-evaluation-scorecard.md.
What belongs here: Sanitized/de-identified datasets (datasets/), executable test cases (cases/), scoring rubrics (rubrics/), and run reports (reports/).
What must not be stored here: Real candidate/employee data in any form — every dataset must be fake/demo or properly de-identified before it lands here.
Owner: [TENANT_CONFIGURATION_REQUIRED — AI Governance Lead], QA co-owns test execution.
Main dependencies: prompts/evaluation/, prompts/red-team/, docs/06-ai-agents-rag/ai-evaluation-strategy.md, .github/workflows/ci.yml.
Subfolders: datasets/ · cases/ · rubrics/ · reports/
Status: Initial scaffold — structure defined, no datasets populated yet.