geo

Warn

Audited by Snyk on Feb 27, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly fetches and analyzes public website content ("Phase 1: Fetch homepage HTML (curl or WebFetch)" and "Extract key pages from sitemap.xml or internal links") and also scans user-generated platforms for brand signals ("Mentions on Reddit, YouTube, Wikipedia, LinkedIn"), meaning untrusted third‑party content is ingested and used to drive scores and recommendations.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 27, 2026, 01:01 PM