The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill installs the @membranehq/cli package from the official NPM registry, which is the expected tool for the author's platform integration.
[COMMAND_EXECUTION]: The skill uses the membrane CLI to execute various webinar management actions, such as creating registrants and listing sessions.
[CREDENTIALS_UNSAFE]: The instructions explicitly follow best practices by directing the agent to use the platform's built-in authentication flow rather than requesting or storing raw API keys or tokens.
[PROMPT_INJECTION]: The skill processes user-provided descriptions and intents to search for or create platform actions. While this presents an indirect prompt injection surface, it is a standard feature of the tool's discovery mechanism.
Ingestion points: SKILL.md (parameters for membrane action list --intent and membrane action create)
Boundary markers: Absent in command-line interpolation examples
Capability inventory: Shell command execution via membrane action run
Sanitization: Relies on the platform's backend for action generation and validation

goto-webinar