The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute membrane CLI commands to manage connections and run actions. These commands are standard for the platform and are used to interact with the Fanout service as intended.
[EXTERNAL_DOWNLOADS]: The skill requires the global installation of the @membranehq/cli package from the NPM registry. This package is the official tool provided by the platform vendor and is a legitimate dependency for the skill's functionality.
[CREDENTIALS_UNSAFE]: The skill demonstrates positive security practices by using the membrane login and membrane connect workflows. It explicitly advises against requesting API keys from the user, relying instead on the platform's server-side credential management.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface through data ingestion via membrane action run and the description-based action creation in membrane action create. However, these are core features of the platform's automation model.
Ingestion points: Output of membrane action run and the description input for membrane action create.
Boundary markers: None explicitly defined in the instructions.
Capability inventory: Command execution using the membrane CLI.
Sanitization: No specific sanitization or validation steps are outlined for the external data returned by actions.

fanout