aidaform
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 ingests and operates on external Aidaform data (user-generated "Form Submissions") via Membrane CLI commands such as the "Running actions" examples and the "Proxy requests" / npx @membranehq/cli request flow in SKILL.md, so arbitrary third‑party form content could be read and influence agent actions.
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 0.90). The skill invokes the external package via "npx @membranehq/cli@latest" at runtime, which fetches and executes remote code (the Membrane CLI) and is required for the skill to operate, so it meets the criteria for a runtime-executed external dependency.
Audit Metadata