Douyin Data Assistant

Problem

Short-video teams often need structured data from public Douyin pages: keyword-based high-engagement video discovery, competitor account posts, comment sentiment, and trending topics. Copying screenshots and manual exports are hard to compare, while browser scraping depends on login state, anti-bot behavior, and brittle formats.

How the skill works

Douyin Data Assistant uses a GUAIKEI_API_TOKEN to fetch public Douyin data through search, author posts, comments, and trending list operations. A typical workflow is: locate candidate videos by keyword with latest or like-count sorting, then fetch public posts from target accounts, comments for a video, or real-time trending items. Results are written to JSON logs, making it easier to sort, deduplicate, aggregate, and build reports with scripts.

• Keyword search: filter by latest or engagement signals for topic selection and competitor content analysis.
• Author posts: batch-fetch public posts from a target account for benchmarking.
• Comments: fetch comment details for a video aweme_id to assess sentiment.
• Trending list: retrieve real-time trending topics for hot-topic tracking.
• Structured export: logs are saved under logs with timestamps, keywords, or IDs in file names for downstream analysis.

Limits

The skill is intended for public, visible data only; it does not access private or hidden content. GUAIKEI_API_TOKEN must be configured before use, and data should remain for personal or internal team analysis rather than unauthorized distribution. If search results are empty, try broader keywords or a wider time range, and respect API quotas and usage restrictions for batch jobs.