uspto-database
Warn
Audited by Snyk on Feb 16, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.70). This skill fetches and ingests public, user-submitted patent and trademark text from third-party USPTO endpoints (e.g., PatentSearch API, PEDS office actions, TSDR, Office Action Text Retrieval and Assignment APIs), which the agent is expected to read and analyze, exposing it to untrusted third-party content that could carry indirect prompt injection.
Audit Metadata