merge

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the @membranehq/cli package from the official NPM registry. This is the documented tool for the Membrane platform and is used for legitimate integration management tasks.
  • [COMMAND_EXECUTION]: The skill guides the user or agent to execute standard CLI commands (membrane login, membrane connect, membrane action) to interact with the service. These operations are scoped to the integration platform's intended use cases and do not involve arbitrary command injection or malicious shell activity.
  • [DATA_EXFILTRATION]: The skill explicitly emphasizes letting the Membrane platform handle credentials server-side, which prevents the need to expose API keys or tokens in the local environment or the agent's context.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 12:51 PM