competitive-analysis

Warn

Audited by Snyk on Feb 18, 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 playbook explicitly instructs the agent to read and synthesize user-generated reviews from open/public third-party sources—e.g., G2, Capterra, Trustpilot, App Store/Google Play, Reddit, and Twitter/X—which are untrusted public content the agent would read and interpret as part of its workflow.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 18, 2026, 01:37 PM