blueshift

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of the @membranehq/cli package from the official npm registry. This is the official tool provided by the vendor (membrane) to facilitate the integration.
  • [COMMAND_EXECUTION]: The skill utilizes the membrane CLI for managing service connections, searching for actions, and executing tasks. These operations are standard for the platform's workflow and do not involve unauthorized privilege escalation or persistence mechanisms.
  • [CREDENTIALS_UNSAFE]: Security best practices are followed as the skill explicitly instructs the agent and user not to handle raw API keys or tokens, instead using the platform's managed connection system.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 03:21 PM
Security Audit — agent-trust-hub — blueshift