Kuaishou Topic Analysis

Problem

Short-video topic planning often gets stuck in vague judgment calls: whether a Kuaishou trend exists, how strong it is, and whether it can be followed manually. Scraping or browsing tends to mix in noise, while content teams need reproducible evidence, not generic direction.

How It Works

The skill reads Kuaishou topic signals through the read-only SocialDataX data layer:
- Retrieves Kuaishou hot lists with ranking, topic, and heat signals
- Searches videos by keyword using kuaishou_search_videos, preserving titles, authors, photo_id, share_url, and visible engagement data
- Supports pagination with page_token, requiring the full next_page_token to be passed back unchanged without truncation, rewriting, or reconstruction
- Separates visible evidence from interpretation, then proposes candidate angles and follow-up questions

A typical workflow starts from a keyword or topic, takes a small sample of hot-list and search results, then narrows the angle using one to three relevant items. If insufficient_balance appears, the skill should stop retrying and surface the recharge link from the error. If pagination breaks, it retains fetched results and suggests changing the input.

Boundaries and Notes

This is a read-only capability. It does not log in, post, like, comment, or read local browser data. It fits content operations, brand research, and creator topic planning, but not content publishing or account automation. For dependency, network, or permission failures, resolve the local environment or authorization first rather than replacing SocialDataX data with public web search.