The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill recommends installing the Membrane CLI package (@membranehq/cli) from the public NPM registry to interact with the service.
[COMMAND_EXECUTION]: The skill instructions utilize the membrane CLI to perform tasks such as user authentication, connection setup, and running various MixMax actions.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill adheres to security best practices by leveraging Membrane's server-side connection management, which avoids the need for hardcoded API keys or manual credential handling by the agent.
[INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes data from external MixMax sources, presenting an attack surface for indirect instructions.
Ingestion points: Data returned from membrane action run commands, including email messages, snippets, and poll results.
Boundary markers: The instructions do not define specific delimiters or guards for the processed data.
Capability inventory: The skill allows the agent to execute CLI commands and dynamically create new actions within the Membrane environment.
Sanitization: There is no mention of sanitization or filtering for the external content fetched from MixMax.

mixmax