skills/affaan-m/ecc/uspto-database/Gen Agent Trust Hub

uspto-database

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill follows secure coding practices by directing users to store API keys in environment variables rather than hardcoding them in the skill or scripts.
  • [EXTERNAL_DOWNLOADS]: The skill facilitates data retrieval from official government sources (uspto.gov) and well-known official-support platforms (patentsview.org).
  • [INDIRECT_PROMPT_INJECTION]: The skill processes data from external USPTO records, creating a theoretical attack surface for indirect prompt injection via record content.
  • Ingestion points: Patent search and trademark record retrieval via API endpoints specified in SKILL.md.
  • Boundary markers: No specific delimiters or boundary markers for external content are defined in the sample code.
  • Capability inventory: The skill is limited to fetching and logging data, with no subprocess or dynamic execution capabilities identified.
  • Sanitization: There is no evidence of specific input sanitization for the data retrieved from the APIs before it is logged or presented to the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 11:08 PM
Security Audit — agent-trust-hub — uspto-database