The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the @membranehq/cli package from NPM. This is an official tool provided by the platform vendor to facilitate integration.
[COMMAND_EXECUTION]: The skill uses the membrane command-line interface to perform authentication, create connections, and execute actions. These commands are part of the intended workflow for interacting with the Membrane ecosystem.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface as it processes external data (action descriptions and schemas) to perform its functions.
Ingestion points: Data enters the agent's context when listing existing actions (membrane action list) or retrieving action details (membrane action get) in SKILL.md.
Boundary markers: The instructions do not define specific delimiters or guardrails to distinguish between trusted instructions and data returned from the CLI.
Capability inventory: The skill has the capability to run arbitrary logic via membrane action run and create new functionality via membrane action create in SKILL.md.
Sanitization: There is no evidence of explicit validation or sanitization of the data retrieved from the API before it is utilized by the agent.

viral-loops