The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user to install @membranehq/cli via npm. This is the official command-line tool provided by the vendor (membranedev) to interact with their platform.
[COMMAND_EXECUTION]: The skill utilizes the membrane CLI to manage integrations. It includes standard commands for authentication (membrane login), connection management (membrane connect), and executing service-specific logic (membrane action run). These commands are necessary for the skill's primary function.
[CREDENTIALS_UNSAFE]: The skill implements secure credential management. It explicitly instructs the agent to never ask for API keys or tokens, delegating the authentication lifecycle and secret storage to the Membrane platform.
[REMOTE_CODE_EXECUTION]: The skill features a capability to create actions from natural language descriptions (membrane action create). While this involves generating code/logic, it occurs within the Membrane platform's managed environment and is a core feature of the service rather than an arbitrary shell execution vulnerability within the agent's local environment.

double