The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill recommends installing the official platform CLI (@membranehq/cli) via NPM. This is a standard practice for the tool's ecosystem and targets a well-known package registry.
[COMMAND_EXECUTION]: The instructions include various membrane CLI commands for authentication, connection management, and performing API actions. These commands are necessary for the skill's functionality and do not exhibit malicious patterns.
[DATA_EXFILTRATION]: While the skill interacts with the Gift Up! API and can read/write data, it does so through the Membrane proxy service. This uses managed authentication headers and is consistent with the skill's primary purpose of integration with a gift card platform.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it ingests data from an external API (Gift Up!).
Ingestion points: Data retrieved from Gift Up! API (e.g., customer names, gift card details, order notes).
Boundary markers: None identified in the instructional content.
Capability inventory: The skill uses the membrane CLI to execute network requests and administrative actions.
Sanitization: No explicit sanitization or validation of the ingested data is described.

gift-up