ki-act-gpai-copyright
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill provides structured instructions for legal analysis of AI models (GPAI) and copyright law compliance. It defines clear professional roles, evidentiary requirements, and formal output formats such as compliance memos and risk matrices.
- [NO_CODE]: The skill is purely informational and instructional. It contains no executable scripts, shell commands, or dynamic code generation. It relies entirely on the agent's reasoning capabilities within the provided legal framework.
- [SAFE]: While the skill mentions external legal databases (such as dejure.org or eur-lex.europa.eu) as authoritative sources for the agent to verify law, it does not contain automated web-scraping code or direct network exfiltration patterns. It instructs the agent to use standard tools to verify citations.
- [SAFE]: There are no indicators of prompt injection, obfuscation, or persistence mechanisms. The metadata and content are consistent with its stated purpose of assisting with EU AI Act and copyright regulatory tasks.
Audit Metadata