analyzing-customers

Warn

Audited by Snyk on Feb 24, 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 instructs the agent to use WebSearch to collect and mine user-generated reviews and social posts from public sources such as G2, Capterra, TrustRadius, app stores, Amazon, Reddit, Twitter, LinkedIn and public forums and to extract quotes/themes that directly drive analysis and recommendations, so untrusted third-party content is read and can influence decisions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 24, 2026, 02:22 AM