Unified Search

Warn

Audited by Snyk on Apr 8, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 1.00). This skill performs live searches against public providers (Tavily, Brave, Linkup, SearXNG) and ingests their search results — explicitly including user-generated sources like Reddit and Hacker News via SearXNG as shown in the SKILL.md uber_search usage and categories — which the agent is expected to read and act on, allowing third-party content to influence behavior.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 8, 2026, 05:26 PM
Issues
1
Security Audit — snyk — Unified Search