The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill installs the official Membrane CLI (@membranehq/cli) from the npm registry. This package is an official tool provided by the vendor for managing integrations.
[COMMAND_EXECUTION]: The skill utilizes shell commands such as membrane login and membrane action run to interface with the Sellercloud platform. These commands are standard operations for the tool's intended functionality.
[CREDENTIALS_UNSAFE]: The instructions follow security best practices by advising the agent never to request API keys or tokens from the user. Authentication is handled via a secure server-side connection model managed by the Membrane platform.
[PROMPT_INJECTION]: The skill processes external data from Sellercloud (e.g., product details, orders), which constitutes an indirect prompt injection surface.
Ingestion points: Data returned by membrane action list and membrane action run commands in SKILL.md.
Boundary markers: No specific delimiters or instructions are provided to the agent to distinguish between its own instructions and data retrieved from Sellercloud.
Capability inventory: The agent can execute and create actions via the Membrane CLI, which are sensitive operations that process this external data.
Sanitization: No explicit sanitization or validation of the Sellercloud data is performed within the skill instructions.

sellercloud