The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill installs the official @membranehq/cli package from the NPM registry. This is the vendor's own command-line tool used to manage the integration.
[COMMAND_EXECUTION]: Shell commands are used to invoke the membrane CLI for authentication (membrane login), connection management (membrane connect), and running integration actions. This is the intended method for interacting with the platform.
[REMOTE_CODE_EXECUTION]: The membrane action create command allows the platform to generate and execute integration logic based on a description. This dynamic code execution is a core feature of the Membrane service and happens within the vendor's managed infrastructure.
[PROMPT_INJECTION]: The skill processes external input through intent-based searching (--intent) and action parameters (--input). While this presents a surface for indirect prompt injection (Category 8), the behavior is necessary for the skill's primary function. Evidence: Ingestion points include command arguments in SKILL.md; boundary markers are absent in examples; capabilities include action execution via the CLI; sanitization is handled by the vendor service.

hologram