fountain

Warn

Audited by Snyk on Apr 21, 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 explicitly instructs using Membrane to proxy requests to the Fountain API (e.g., "membrane request CONNECTION_ID /path/to/endpoint" and "membrane action run"), which fetches user-generated/third-party Fountain content that the agent would read and act on.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 21, 2026, 02:29 PM
Issues
1