circle-1

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to install the official Membrane CLI tool (@membranehq/cli) from the public NPM registry in SKILL.md to enable communication with the platform.
  • [COMMAND_EXECUTION]: Utilizes the membrane CLI in SKILL.md to perform operations such as user authentication, connection management, and running community-related actions.
  • [REMOTE_CODE_EXECUTION]: Supports the creation and execution of dynamic logic on the Membrane platform through natural language intent processing via the membrane action create and membrane action run commands described in SKILL.md.
  • [DATA_EXFILTRATION]: The skill utilizes a managed platform (Membrane) to handle the full OAuth lifecycle server-side, ensuring that no sensitive credentials or API keys are stored or handled locally by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 05:37 PM
Security Audit — agent-trust-hub — circle-1