idea-os

Warn

Audited by Snyk on Apr 28, 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's Phase 3 "Research" step explicitly requires WebSearch + WebFetch (minimum 5 WebSearches and 2 WebFetches on named competitors, sourcing TAM numbers and community signals like subreddits), meaning the agent will fetch and read untrusted public web content that directly influences the PRD and plan.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 28, 2026, 08:49 PM
Issues
1