The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user to install the official Membrane CLI package (@membranehq/cli) from the NPM registry to enable its functionality.
[COMMAND_EXECUTION]: The skill uses various shell commands, including npm install for initial setup and membrane commands for authentication, connection management, and running package management actions.
[PROMPT_INJECTION]: The skill exposes an indirect prompt injection surface where natural language intent provided by the user is used to search for or create actions.
Ingestion points: External data enters the context via the QUERY and DESCRIPTION placeholders used in the membrane action list and membrane action create commands within SKILL.md.
Boundary markers: No boundary markers or instructions to ignore embedded commands are present in the command templates for the interpolated natural language strings.
Capability inventory: The skill is capable of executing actions via membrane action run, which can modify system state or interact with the Chocolatey package manager.
Sanitization: The instructions do not describe any sanitization, escaping, or validation of the user-provided natural language before it is passed to the CLI tools.

chocolatey