rasa

Pass

Audited by Gen Agent Trust Hub on May 1, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the @membranehq/cli package from the NPM registry, which is the official tool provided by the vendor for managing integrations.
  • [COMMAND_EXECUTION]: Uses various membrane CLI commands to perform administrative and operational tasks such as logging in, listing connections, and searching for actions.
  • [REMOTE_CODE_EXECUTION]: Provides functionality to dynamically create and run actions via the membrane action create and membrane action run commands. This logic is executed on the Membrane platform as part of the primary purpose of the skill.
  • [CREDENTIALS_UNSAFE]: Explicitly instructs the agent and user to avoid local secret management, directing them to use the Membrane platform's built-in OAuth and connection handling to manage credentials securely.
Audit Metadata
Risk Level
SAFE
Analyzed
May 1, 2026, 01:15 AM
Security Audit — agent-trust-hub — rasa