NYC

User Researcher

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: 0.70). The skill explicitly expects collecting and analyzing untrusted user-generated content—e.g., recruiting from social media/communities ("LinkedIn, Reddit, Slack") and ingesting participant data such as "Surveys (open-ended)", "All sessions recorded (with permission)", "transcripts" and "artifacts collected (screenshots, workflows)"—which the agent would read/interpret as part of synthesis and reporting.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 15, 2026, 08:46 PM