golden-dataset-validation
Pass
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements robust data validation logic, including JSON schema enforcement and semantic similarity checks using standard libraries like NumPy.- [INDIRECT_PROMPT_INJECTION]: The skill processes external dataset content (documents and queries). However, it employs strict schema validation and quality rules. The attack surface is minimal as the data is used for validation and statistical analysis rather than control flow or command generation.- [COMMAND_EXECUTION]: The skill documents CLI integration and pre-commit hooks using
uv run. These represent standard development workflows for the intended Python environment and are used for running local validation scripts.
Audit Metadata