The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user to install a global Node.js package (@membranehq/cli@latest) from the NPM registry to enable its functionality.
[COMMAND_EXECUTION]: The skill relies extensively on executing shell commands via the membrane CLI tool, including logging in, connecting to services, and running actions.
[COMMAND_EXECUTION]: The membrane action create command implements dynamic execution by generating and building new code logic (actions) at runtime based on natural language descriptions.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes external weather data from the AerisWeather API and takes user-provided descriptions to generate executable actions.
Ingestion points: Weather data from AerisWeather APIs; intent strings in membrane action list; and descriptions in membrane action create (SKILL.md).
Boundary markers: The skill suggests using the --json flag to enforce structured data boundaries for machine readability.
Capability inventory: Subprocess execution of the membrane CLI and network operations via API connectors (SKILL.md).
Sanitization: The skill delegates authentication and connection management to the Membrane platform but does not explicitly define sanitization for the content of generated actions.

aerisweather