The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the membrane CLI to manage connections and execute actions on the Membrane platform.
Evidence: Commands like membrane action run, membrane action list, and membrane action create are used to interact with the service.
[EXTERNAL_DOWNLOADS]: The instructions specify installing the @membranehq/cli package from the NPM registry.
Evidence: npm install -g @membranehq/cli@latest and npx @membranehq/cli@latest are used to set up the necessary tooling.
[DATA_EXFILTRATION]: The skill implements a security-positive pattern by utilizing server-side credential management.
Evidence: The 'Best practices' section explicitly instructs the agent to 'never ask the user for API keys or tokens' and to 'Let Membrane handle credentials'.
[PROMPT_INJECTION]: As the skill processes external data from Discord, there is a theoretical surface for indirect prompt injection.
Evidence: The skill processes output from Discord actions which could contain malicious instructions, though it lacks specific boundary markers in the current instructions.

discord