The Agent Skills Directory

[SAFE]: The skill includes two Python scripts, contract_validator.py and risk_tier_classifier.py, which perform local static analysis on JSON input. These scripts rely exclusively on the Python standard library and do not perform network operations, execute shell commands, or access sensitive system files.
[SAFE]: External references in SKILL.md and data/sources.json target official regulatory bodies (European Union AI Act), standards organizations (NIST), and well-known technology providers (OpenAI, Anthropic, Linux Foundation) for documentation and compliance guidance.
[SAFE]: Analysis of the skill instructions and metadata revealed no evidence of prompt injection, data exfiltration, or persistence mechanisms. The YAML frontmatter is consistent with the skill's stated purpose.

ai-product-operating-model