The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the @membranehq/cli package via NPM. This is an official utility provided by the vendor (Membrane) to manage platform interactions and authentication.
[COMMAND_EXECUTION]: The instructions rely on the execution of various membrane CLI commands to perform authentication, manage connections, and execute data actions within the Anodot environment.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting data from the Anodot API that is subsequently processed by the AI agent.
Ingestion points: External data is retrieved through the membrane action run command and returned in the output field.
Boundary markers: Absent. There are no explicit instructions to the agent to treat the retrieved data as untrusted or to use specific delimiters.
Capability inventory: The agent can search for, create, and execute arbitrary actions via the CLI, which includes reading and writing data to the Anodot platform.
Sanitization: Not specified. The skill does not provide mechanisms for validating or sanitizing the content returned from external API calls before it enters the agent's context.

anodot