mumara

Pass

Audited by Gen Agent Trust Hub on Apr 21, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the official @membranehq/cli package from NPM, which is a trusted tool provided by the vendor (Membrane) for platform interactions. This is a standard and safe procedure.\n- [COMMAND_EXECUTION]: It utilizes the membrane CLI to perform necessary integration tasks, including authentication (membrane login), connector discovery (membrane search), and executing API actions (membrane action run). These commands are the intended way to interact with the platform.\n- [DATA_EXFILTRATION]: API requests are routed through a proxy (membrane request) that handles authentication server-side. This design prevents raw API keys or sensitive credentials from being exposed in the local environment or the agent's context.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 21, 2026, 11:19 PM