Use when a user starts a task in natural language and the agent may need additional structured information before proceeding. This skill guides the agent to decide whether intake is needed, generate an external self-contained form link/page, ask the user to fill it, parse the copied JSON submission, and continue the task. Works without native frontend card rendering or a backend by using copy-back form submissions.
Use when a user starts a task in natural language and the agent may need additional structured information before proceeding. This skill guides the agent to decide whether intake is needed, generate an external self-contained form link/page, ask the user to fill it, parse the copied JSON submission, and continue the task. Works without native frontend card rendering or a backend by using copy-back form submissions.