setup

Warn

Audited by Snyk on Feb 20, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.80). The skill's setup explicitly includes external research tooling ("Research tool detection" in Phase 5 and ops/config.yaml) and generates a /learn skill (Step 9) that uses Exa MCP or "web-search" as a last resort, indicating the agent will fetch and ingest public web content which it will read and act on.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 20, 2026, 02:18 AM