The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on the membrane CLI to perform operations such as managing connections, discovering available actions, and executing tasks on the Chatfuel platform. This is a standard and secure method for delegating API interactions to a managed service.
[EXTERNAL_DOWNLOADS]: The skill requires the installation of the @membranehq/cli package from the official npm registry. This package is the legitimate CLI tool provided by the vendor (membranedev) for their platform.
[PROMPT_INJECTION]: The skill demonstrates a surface for indirect prompt injection by processing natural language strings for action discovery and creation. This is an inherent feature of the platform's functionality.
Ingestion points: Untrusted data enters the context via the --intent "QUERY" and membrane action create "DESCRIPTION" arguments in SKILL.md.
Boundary markers: No explicit boundary markers or delimiters are defined in the command examples.
Capability inventory: The skill provides the ability to execute generated actions via membrane action run as documented in SKILL.md.
Sanitization: No client-side sanitization is performed in the instructions; the skill relies on the Membrane platform to handle validation and safe execution.

chatfuel