audience-profiler

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's required workflow (Steps 3 and 4) uses Perplexity tools (perplexity_ask / perplexity_search) to fetch and cite community discussions, reviews, specific online communities (e.g., Reddit, Slack/Discord groups, Facebook groups), and public media sources, meaning the agent will ingest and act on untrusted, user-generated third‑party content that can influence downstream profiling and actions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 27, 2026, 01:00 AM