The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the installation of the @membranehq/cli package from npm. This is an official tool provided by the vendor to manage integrations and authentication securely.
[COMMAND_EXECUTION]: The skill uses the membrane command-line interface to perform actions like authentication, connection management, and running integration logic. These commands are scoped to the intended functionality of the skill.
[CREDENTIALS_SAFE]: The skill explicitly advises against handling raw API keys or tokens, directing users to use its built-in connection management system. This reduces the risk of credential exposure in logs or environment variables.
[INDIRECT_PROMPT_INJECTION]: The skill accepts natural language descriptions to search for or create actions. While this is a data ingestion surface, the risk is limited as the input is used as a query for the vendor's platform rather than being directly executed as code.

userflow-1