The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill installs the @membranehq/cli tool globally using npm. This is the official command-line interface provided by the vendor, membranedev, to manage integrations and is a standard dependency for this skill's functionality.\n- [COMMAND_EXECUTION]: Instructions rely on the membrane CLI to perform operations such as authentication, action discovery, and task execution. This centralizes interaction through a purpose-built tool.\n- [DATA_EXFILTRATION]: The skill handles sensitive information like contact details and business leads. It correctly utilizes a managed connection service that prevents direct exposure or local storage of API keys and tokens.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from Apex 27 which could potentially contain malicious instructions intended to influence the agent. Evidence chain: data enters via membrane action run output; no explicit delimiters or boundary markers are documented; the agent has shell execution capabilities via the CLI; no specific sanitization steps are provided in the skill instructions.

apex-27