social-media-trends-research
Warn
Audited by Snyk on Feb 15, 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). This skill programmatically fetches and ingests open/public third‑party content—notably Reddit posts via public .json endpoints (scripts/reddit_scraper.py), Google Trends via pytrends, and Perplexity MCP queries of Twitter/TikTok/Web—and the agent reads and interprets that user-generated/untrusted content to produce recommendations, creating a risk of indirect prompt injection.
Audit Metadata