social-intelligence

Warn

Audited by Snyk on Feb 19, 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). The SKILL.md explicitly describes searching and ingesting public, user-generated social media posts (e.g., "Search posts across Twitter, Instagram, and Reddit" and the "35 MCP Tools Available" list), so the agent will read and act on untrusted third‑party content fetched from social platforms.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 19, 2026, 01:51 AM