Douyin Works Insight

Problem Addressed

When reviewing a Douyin account, the friction is often not finding content, but assembling scattered signals: follower count, recent works, likes, comments, shares, links, and content patterns are hard to compare manually. For competitor monitoring, creator evaluation, or viral-content breakdown, copying pages into spreadsheets is slow and easy to miss key posts. Douyin Works Insight targets this workflow: given an account nickname or ID, it returns a structured content snapshot instead of raw page fragments.

How It Works

The skill uses the Redfox Data API and authenticates with REDFOX_API_KEY, without requiring a Douyin login. The core flow is centered on DouyinWorksFetcher:

• Account recognition: automatically treats input as a nickname or Douyin ID, reducing manual mode switching.
• Basic profile retrieval: extracts follower count, likes, total works, Redfox index, and other account-level metrics.
• Recent works extraction: fetches up to 50 recent works in reverse chronological order, including likes, comments, shares, interaction counts, and work links.
• Highlight analysis: analyzes the top 3 works by interaction and adds account-level observations such as update frequency, engagement performance, and viral-content patterns.
• Formatted output: supports Markdown for reading and JSON for downstream analysis or reporting pipelines.

Boundaries and Caveats

It is best treated as an account snapshot and recent-content analyzer, not a full historical database. The works list is capped at 50 entries, while the historical total may appear in awemeCount. Nicknames can be ambiguous, so precise queries should prefer Douyin IDs. The API is usage-based, and failures may come from quota, network timeouts, or missing data; retries and key status should be checked before drawing conclusions. For reports, use the output as structured input rather than a replacement for editorial judgment.