alchemer
Warn
Audited by Snyk on Mar 4, 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). This skill explicitly retrieves data from Alchemer (e.g., "List Survey Responses" / "get-survey-response" in SKILL.md and via the Membrane proxy command
membrane request CONNECTION_ID /path/to/endpoint), which exposes the agent to untrusted, user-generated survey content that the agent is expected to read and could materially influence subsequent actions.
Audit Metadata