superdocu

Pass

Audited by Gen Agent Trust Hub on Apr 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references the installation of the @membranehq/cli tool, which is an official utility for the platform.
  • [PROMPT_INJECTION]: The skill handles data from the Superdocu API, which introduces a potential surface for indirect prompt injection. Ingestion points: Data retrieved from the Superdocu API via membrane CLI commands in SKILL.md. Boundary markers: Not defined in the current instructions. Capability inventory: Ability to read and write to the Superdocu API using the membrane CLI. Sanitization: Not explicitly handled in the skill instructions.
  • [SAFE]: The skill utilizes a secure proxy and authentication management system provided by the Membrane platform, adhering to security best practices for credential handling.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 22, 2026, 08:52 PM