The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes external data from Square payments, orders, and invoices that could potentially contain malicious instructions targeting the AI agent.
Ingestion points: Untrusted data enters the agent context through tools like SQUARE_LIST_PAYMENTS, SQUARE_SEARCH_ORDERS, and SQUARE_LIST_INVOICES.
Boundary markers: The skill lacks explicit instructions or delimiters to ensure the agent ignores potential commands embedded within retrieved data fields.
Capability inventory: The agent possesses significant capabilities including the ability to cancel payments, update orders, and cancel invoices.
Sanitization: No sanitization or validation of the retrieved Square data is described in the skill.
[EXTERNAL_DOWNLOADS]: The skill requires the user to connect to an external, third-party MCP server endpoint at https://rube.app/mcp for its operation.

square-automation