The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches and installs the @membranehq/cli package from the official NPM registry. This tool is the primary interface for the Membrane platform and is a vendor-provided resource.
[COMMAND_EXECUTION]: Executes various membrane CLI commands to perform authentication, manage MLflow connections, and trigger remote actions. These commands are part of the intended functionality for interacting with the Membrane ecosystem.
[PRIVILEGE_ESCALATION]: The instructions recommend a global installation of the CLI tool (npm install -g), which typically requires administrative or elevated system privileges on most operating systems.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external MLflow experiments, runs, and models. This creates a surface for potential indirect prompt injection if the source data contains malicious instructions.
Ingestion points: Data retrieved from MLflow API endpoints via membrane action run and membrane request commands.
Boundary markers: No specific delimiters or instructions to ignore embedded content are present in the provided skill documentation.
Capability inventory: The skill utilizes the membrane CLI to perform network operations, manage system-wide authentication states, and execute remote functions.
Sanitization: The instructions do not specify any validation, filtering, or escaping mechanisms for the data retrieved from MLflow before it is processed by the agent.

mlflow