About this skill

Problem Solved

When engineers need to wire Douyin hot-list and work-search data into automated workflows, scraping browsers or hand-rolling endpoints can introduce login state, anti-bot issues, and unclear permission boundaries. This skill wraps read-only data calls for trend scanning, content planning, and competitor observation.

How It Works

Prefer direct CLI: when shell commands are available, use the direct CLI without MCP server configuration. Hot-search queries require no arguments; video/work search requires --keyword.
Optional filters: supports --sort-type, --publish-time-range, --duration-range, --content-type, and --page-token pagination; --pretty only changes formatting.
MCP compatibility: if MCP tools are present, call douyin_get_hot_search_list for the hot list and douyin_search_videos for keyword search.
Output discipline: separate observed evidence from interpretation, and preserve traceable fields such as rank, keyword, heat, content IDs, URLs, authors, publish time, and content_type.

Boundaries

The skill is read-only: it does not read local browser data, save API keys, log in, post, like, comment, or change account state. The API key environment variable is fixed as SOCIALDATAX_API_KEY; key acquisition or management should use the official website entry only, without inferring alternate domains. Continue pagination only when next_page_token is non-empty; do not stop merely because one page has empty items.

Screenshots

Use Cases

Before a content meeting, use Douyin hot lists and keyword search to find high-heat topics.
For competitor analysis, search works by keyword, publish time, and content type, then keep traceable fields.
In market observation scripts, call the hot-list API to get rank, keyword, heat, and linked content.
In MCP environments, paginate douyin_search_videos to collect multiple pages of work samples.

Best For

Data analysts: connect Douyin hot lists and work search into topic or competitor analysis.
Content operations: scan hot keywords, high-heat works, and publish time before planning.
AI engineers: call read-only Douyin search and hot-list tools in agents or MCP.
Market researchers: collect titles, authors, URLs, publish time, and content_type.