insided

Warn

Audited by Snyk on Apr 22, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.80). The SKILL.md instructs using Membrane to connect to the public InSided API and run proxy requests or actions that read Articles, Posts, and Comments (user-generated community content), so the agent will fetch and interpret untrusted third-party content from InSided which could influence subsequent actions.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 22, 2026, 04:39 PM
Issues
1