tikhub-social-media（Tikhub 社交媒体）

TikHub Social Media Data Query

Setup (Required)

Set your API key before using this skill:

export TIKHUB_API_KEY="<your-tikhub-api-key>"

Get your API key at: https://tikhub.io (register → dashboard → API Keys)

Optional proxy (if TikHub API is blocked in your network):

export TIKHUB_PROXY="http://your-proxy:port"

Authentication

• Base URL: https://api.tikhub.io
• Auth Header: Authorization: Bearer <TIKHUB_API_KEY>
• All responses: unified ResponseModel → {code, message, data, cache_url, ...}
• Check balance: GET /api/v1/tikhub/user/get_user_info

Helper Script

Use scripts/tikhub_query.py for all API calls:

python3 scripts/tikhub_query.py <endpoint_path> [param=value ...] [--method POST] [--body '{}']

Examples:

# Search Douyin videos
python3 scripts/tikhub_query.py /api/v1/douyin/web/fetch_search keyword=宁德时代

# Get TikTok user profile
python3 scripts/tikhub_query.py /api/v1/tiktok/web/fetch_user_profile secUid=MS4wLjABAAAA...

# Search Xiaohongshu notes
python3 scripts/tikhub_query.py /api/v1/xiaohongshu/web/get_note_by_keyword keyword=CATL

# Get Instagram user info
python3 scripts/tikhub_query.py /api/v1/instagram/v1/fetch_user_info username=cataborrecycling

# Search Twitter
python3 scripts/tikhub_query.py /api/v1/twitter/web/fetch_search_timeline keyword="CATL battery"

# Search YouTube
python3 scripts/tikhub_query.py /api/v1/youtube/web/search_videos keyword="CATL solid state battery"

# Get Weibo hot search
python3 scripts/tikhub_query.py /api/v1/weibo/web/fetch_hot_search

# Hybrid parse any video URL
python3 scripts/tikhub_query.py /api/v1/hybrid/video_data url=https://v.douyin.com/xxx

Platform Quick Reference

Endpoint Naming Convention

All endpoints follow: /api/v1/{platform}/{source}/{action}

• platform: douyin, tiktok, xiaohongshu, instagram, youtube, twitter, weibo, kuaishou, bilibili, etc.
• source: web, app, app/v3, web_v2, etc.
• action: fetch_user_profile, fetch_one_video, fetch_search, etc.

Core Actions (available on most platforms)

Platform → Preferred API Version

Pagination

Most list endpoints support cursor-based pagination:

• cursor / max_cursor / offset / after — pass from previous response
• count / page_size — items per page (typically 10-30)
• Response includes has_more, cursor, next_cursor etc.

Hybrid Video Parser

Parse any video URL from supported platforms automatically:

GET /api/v1/hybrid/video_data?url=<any_video_url>

Supports: Douyin, TikTok, Xiaohongshu, Kuaishou, Bilibili, Weibo, etc.

Detailed Platform References

For full endpoint lists per platform, read the corresponding reference file:

• references/douyin.md — Douyin (Web + App + Search + Billboard + Creator + Xingtu)
• references/tiktok.md — TikTok (Web + App + Creator + Analytics + Ads + Shop)
• references/xiaohongshu.md — Xiaohongshu (App V2 + Web V2)
• references/instagram.md — Instagram (V1 + V2 + V3)
• references/youtube.md — YouTube (Web + V2)
• references/twitter-threads.md — Twitter/X + Threads
• references/reddit-linkedin.md — Reddit + LinkedIn
• references/chinese-platforms.md — Weibo, Kuaishou, Bilibili, WeChat, Zhihu, Toutiao, etc.

Only read the relevant reference file when you need endpoint details for a specific platform.

Cost Awareness

• Most endpoints: $0.001-$0.005 per call
• Some premium endpoints (Instagram, creator analytics): up to $0.01
• Always prefer cached results when cache_url is returned
• Batch operations where available (e.g., Reddit batch post details)
• Check balance before bulk operations: GET /api/v1/tikhub/user/get_user_info