Douyin Keyword Work Search

What It Solves

When analyzing short-video content, a broad keyword often produces noisy results. Terms like “fashion” or “food” can mix many unrelated creators and formats, while a single viral post lacks enough context to judge what is actually resonating. Douyin Keyword Work Search targets this workflow: use a keyword to surface high-interaction Douyin works and compare them by structured metrics such as likes, comments, shares, and collections.

How It Works

The skill breaks the query into explicit steps:
- Keyword extraction: it prioritizes refined terms such as “niche film” or “workplace outfit” instead of broad categories, and strips trigger words like “search,” “viral,” and “trending works.”
- Broad-term expansion: when a keyword is too general, it proposes about 10 narrower candidate terms and waits for confirmation before querying.
- Time parsing: it supports absolute dates, relative dates, and ranges, such as “last 7 days,” “June 1,” or “May 30 to June 2”; vague expressions leave the time filter unset.
- Structured results: the response includes title, author, like_count, comment_count, share_count, collect_count, work_url, and publish_time, organized for comparison.

Boundaries

It is most useful for trend scanning, competitor content monitoring, and keyword-level content discovery. It depends on a configured REDFOX_API_KEY, and results are constrained by the keyword, time window, and data returned by the platform; zero matches only indicate a need for expanded search terms, not that no relevant content exists.