The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the @membranehq/cli package from the NPM registry to function. This is a vendor-provided tool.
[COMMAND_EXECUTION]: Instructions involve running several membrane CLI commands to manage authentication, list connections, and execute actions.
[REMOTE_CODE_EXECUTION]: The skill enables the execution of remote logic via membrane action run and the dynamic generation of new logic through membrane action create.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by processing untrusted data from external sources (MoDeck/Twitter).
Ingestion points: Output from membrane action list and membrane action run are ingested into the agent context as described in SKILL.md.
Boundary markers: No delimiters or specific isolation instructions are defined for processing the data returned by the CLI.
Capability inventory: The skill has access to local shell execution via the membrane CLI.
Sanitization: No specific sanitization, validation, or filtering of the API response data is described before it is processed by the agent.

modeck