patent-claim-analyzer

Warn

Audited by Snyk on Jun 28, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.95). The skill calls outsider-authored free text from the public patent-claims database via runtime API responses (Zhihuiya endpoints claimData / claimDataTranslated / bibliography), and then injects that returned claim text into the LLM context as part of the generated markdown report.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 28, 2026, 07:06 AM
Issues
1
Security Audit — snyk — patent-claim-analyzer