epc-search

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions and metadata are entirely consistent with the stated purpose of patent law research. No deceptive metadata or malicious intent was found.\n- [SAFE]: The skill uses standard tools (Bash, Read, Write) for the intended purpose of executing search queries and writing legal memoranda. No misuse of these tools was observed.\n- [PROMPT_INJECTION]: The skill has a data ingestion surface (Category 8) for legal search results, but it is not considered a threat. Evidence chain:\n
  • Ingestion points: Legal data returned by search_patent_law and search_mpep tools (SKILL.md).\n
  • Boundary markers: The skill uses clear markdown headers and thematic separators (e.g., '---') to delimit research results.\n
  • Capability inventory: Bash, Read, and Write tools for processing and saving data.\n
  • Sanitization: Legal provisions from authoritative sources (EPO, WIPO) are considered low-risk, static content.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 01:28 PM
Security Audit — agent-trust-hub — epc-search