customer-research

Warn

Audited by Snyk on Aug 26, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (medium risk: 0.30). The skill’s “Mode 2: Go Find Research” runtime workflow explicitly directs reading outsider-authored content from online communities/review sites (e.g., Reddit/G2) and extracting verbatim quotes from those sources without requiring the agent to pre-select trusted items first.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 26, 2026, 05:56 PM
Issues
1
Security Audit — snyk — customer-research