dacast

Pass

Audited by Gen Agent Trust Hub on May 3, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill guides the user to install the official Membrane CLI (@membranehq/cli) via the npm registry. This is a standard installation of a vendor-provided tool required for the skill's functionality.
  • [COMMAND_EXECUTION]: The skill utilizes the membrane CLI to perform video management tasks, such as searching for actions and executing API calls. These commands are used as intended for the integration and do not perform unauthorized system modifications.
  • [CREDENTIALS_UNSAFE]: The skill explicitly advises against asking users for API keys or tokens. Instead, it uses a secure authentication flow (membrane login and membrane connect) that manages credentials server-side, reducing the risk of secret exposure in the agent context.
Audit Metadata
Risk Level
SAFE
Analyzed
May 3, 2026, 04:47 AM
Security Audit — agent-trust-hub — dacast