Xingtu Task Author Fetcher

Problem

Xingtu recruitment tasks often return registered creators across paginated API responses, which makes manual copying prone to missing fields or inconsistent ordering. This skill turns “fetch all registered authors for a task and save them to a spreadsheet” into a repeatable workflow, useful when the creator list needs to be handed off to business, operations, or campaign teams.

Workflow and output

• Task identification: Prefer the Xingtu task ID in the message; if the user supplies a URL or screenshot, extract the long numeric ID from it.
• Login state handling: Check ~/.xingtuCookie.txt, prompt for login when needed, and validate the Cookie to reduce failed API calls.
• Paginated fetching: Call the Xingtu task author list endpoint and page through results until pagination.has_more is false.
• Excel export: Normalize fields such as creator nickname, author ID, follower count, pricing, expected views, completion rate, enrollment status, and recommendation reason, then output an .xlsx file.
• Result delivery: Show the summary count and deliver the Excel file to the user.

Boundaries

The skill depends on a valid Xingtu backend Cookie and accessible Xingtu API. If the Cookie expires, the network fails, or the API returns an empty list, it should prompt for re-login or report the error. It is scoped to exporting the registered-author list for a task; it does not create tasks, edit creator records, or run downstream business processes.