yuja

Pass

Audited by Gen Agent Trust Hub on Apr 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the @membranehq/cli package from the official npm registry. This is a vendor-owned resource for the 'membranedev' author and is used to manage integrations.
  • [COMMAND_EXECUTION]: The skill utilizes the membrane command-line tool to perform operations such as authentication, listing connections, and executing actions against the YuJa API. These operations are within the scope of the skill's intended functionality.
  • [DATA_EXFILTRATION]: Provides functionality to send HTTP requests to the YuJa API via the membrane request command. This uses the vendor's infrastructure to securely manage authentication headers and base URLs.
  • [PROMPT_INJECTION]: The skill instructions emphasize using pre-built actions and letting the platform handle credentials, which reduces the surface area for common injection or credential theft attacks.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 22, 2026, 07:40 AM