The Agent Skills Directory

[SAFE]: The skill provides instructions for installing and using the @membranehq/cli tool. This is a legitimate tool provided by the vendor (Membrane) and is installed from the official NPM registry.
[DATA_EXFILTRATION]: The instructions emphasize using Membrane's server-side connection management rather than handling sensitive API keys locally, which is a recommended security practice to prevent credential exposure.
[COMMAND_EXECUTION]: The skill uses standard CLI commands to perform its functions. These operations are transparent and consistent with the skill's purpose as a developer tool.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external API outputs.
Ingestion points: Data returned by membrane action run and membrane action list (SKILL.md).
Boundary markers: None identified.
Capability inventory: Local command execution via the membrane CLI.
Sanitization: None identified in the provided instructions; behavior depends on the agent's own safety constraints.

carbon-ads