patent

Warn

Audited by Snyk on Jun 14, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (high risk: 0.85). The skill’s runtime LLM context is populated from outsider-authored free text scraped from public patent sources (e.g., Google Patents/Espacenet/USPTO PPS pages and Lens.org citation data) during Phase 3–6, which are not authored by the operating user.

Issues (1)

W011
MEDIUM

Third-party content exposure detected (indirect prompt injection risk).

Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 14, 2026, 02:34 PM
Issues
1
Security Audit — snyk — patent