skills/joker-of-gotham/finkg-evaluation-skills/fin-kg-define-quality-model-requirements/Gen Agent Trust Hub
fin-kg-define-quality-model-requirements
Pass
Audited by Gen Agent Trust Hub on Jul 17, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill contains multiple Python utility scripts (e.g.,
audit_quality_bundle.py,validate_quality_artifacts.py) designed to automate data validation and artifact management. These scripts use standard Python libraries likepathlibandargparseto interact with files in the local environment. No arbitrary or high-risk command execution patterns were found. - [DATA_EXFILTRATION]: There are no network operations or data transmission patterns. All scripts operate purely on local YAML and JSON files within the skill's directory structure.
- [PROMPT_INJECTION]: The
SKILL.mdfile and reference documents contain strict domain-specific instructions (e.g., "Never define measurement formulas", "Strict constraint") intended to enforce a governance boundary. These are legitimate instructional constraints for the AI agent to follow when generating data and do not represent attempts to bypass safety filters. - [REMOTE_CODE_EXECUTION]: The skill does not fetch or execute code from external sources. Its only dependency is the widely-used
PyYAMLlibrary, and it usesyaml.safe_loadfor secure parsing. - [NO_CODE]: Although the skill contains Python code, it is utilized for offline data validation and structure enforcement rather than for dynamic or autonomous system interactions.
Audit Metadata