Douyin Content Research

Problem Solved

When planning Douyin content, operators and creators often need to combine hot-search rankings, related videos or image posts, account names, publish times, and engagement signals from scattered sources. Douyin Content Research is a read-only research workflow for turning keywords or content directions into reviewable sample lists and follow-up angles.

How It Works

The skill prefers direct CLI calls to SocialDataX, with MCP tools available when the agent can already call them. Core inputs are a keyword or a hot-search request, plus optional filters such as sort_type, publish_time_range, duration_range, content_type, pages, and max-items. It uses SOCIALDATAX_API_KEY from the environment and does not require extra MCP server setup.
Results are JSON with data.items, page_count, item_count, and pagination markers. Keep hot-list signals separate from keyword search results, and preserve traceable fields such as title, author, link or content ID, and publish time. Use returned images for image posts, and treat video.media_type="audio" as an audio resource.

Boundaries

This is read-only: it does not log in, publish, like, comment, or modify accounts. If dependencies, network, permissions, or balance errors occur, follow the returned guidance. Broaden keywords or remove filters when results are sparse; add scene, audience, brand, time range, or account names when results are noisy.