dandelion

Pass

Audited by Gen Agent Trust Hub on Apr 21, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the Membrane CLI (@membranehq/cli) for session management and API interaction. These commands facilitate authentication and task execution through the vendor's platform, which is standard behavior for skills authored by this developer.
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the vendor's official command-line interface from the public npm registry and provides links to authorized documentation and source repositories.
  • [DATA_EXFILTRATION]: No unauthorized data transfer patterns were detected. The skill specifically instructs the agent to use Membrane's secure proxy for all external requests, ensuring that credentials and tokens are managed server-side rather than being handled directly by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 21, 2026, 07:44 PM