The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs users to install the official @membranehq/cli package from the npm registry. This is the primary method for interacting with the Membrane platform and originates from a recognized vendor resource.
[COMMAND_EXECUTION]: The skill utilizes several membrane CLI commands to perform essential tasks such as authentication, service connection, and action execution. These commands are part of the intended functionality for managing Remarkety marketing data.
[PROMPT_INJECTION]: This category identifies potential surfaces for indirect prompt injection where untrusted data might influence agent behavior.
Ingestion points: Untrusted data enters the context via the --input parameter in the membrane action run command and the "DESCRIPTION" argument in membrane action create (SKILL.md).
Boundary markers: No explicit boundary markers or delimiters are suggested for the input JSON or the action descriptions.
Capability inventory: The skill has the capability to read and write data to Remarkety (campaigns, contacts, orders) and create new executable actions via the Membrane platform.
Sanitization: The skill relies on the Membrane platform's internal processing to validate action descriptions and input schemas.

remarkety