
China Data: Taobao, Bilibili, WeChat, Trending
io.github.LMSlay由 GitHub 帳號 LMSlay 發布v1.3.1更新於 Oct 26, 2026
Taobao/Tmall, Bilibili videos + transcripts, WeChat articles, China hot search, suppliers by photo.
概覽
託管於 Apify 的遠端 MCP 伺服器,讓助理取得中國市場資料:淘寶/天貓商品、B 站影片與轉錄、微信公眾號文章、熱搜榜,以及依圖片找供應商。
- 可以做什麼
- 搜尋淘寶/天貓商品列表、價格與賣家
- 探索 B 站影片並取得其轉錄文字
- 擷取公開的微信公眾號文章
- 讀取中國熱搜榜單快照
- 依商品照片搜尋阿里巴巴/義烏供應商
- 適用情境
- 適合研究中文資料來源或中國電商平台時使用:查詢淘寶/天貓商品與價格、取得 B 站內容與語音轉錄、閱讀公開微信公眾號文章、掌握目前熱搜話題,或從商品圖片整理供應商候選名單。
- 執行需求
- 使用 Apify 託管 MCP 伺服器上的遠端端點,無需安裝。需要 Apify 帳號並以 OAuth 登入;不支援 OAuth 的用戶端須以 Authorization Bearer 標頭提供 Apify API 權杖。需要能連線至該端點。
安裝
在 SourceWeft 中
- 開啟 儀表板中的 China Data: Taobao, Bilibili, WeChat, Trending,將其新增到工作區。
- 連接 所需的 憑證。它們會被私密儲存,絕不會在公開頁面上展示。
- 為需要使用其工具的對話啟用該服務。
Web executable,透過 Streamable HTTP。 遠端服務在工作區中設定後即可從網頁執行環境執行。
交給代理程式安裝
把這段提示詞貼上到 Claude Code、Codex、Cursor 等代理程式,或 SourceWeft 對話中。代理程式會讀取該伺服器的安裝指南,先向你展示發布者、端點和登入方式,經你同意後再把它新增到你的客戶端。金鑰不會寫進任何命令。
閱讀 https://sourceweft.com/mcp/io.github.LMSlay%2Ftubetext-china-data/install.md,並按其中說明把這個 MCP 伺服器接入我的客戶端。修改任何設定之前,先告訴我它的發布者、端點和登入方式,等我確認。不要把 API 金鑰寫進命令或我會提交的檔案;問我金鑰應從哪裡取得。其他 MCP 客戶端
把它新增到你客戶端的 mcpServers 設定中。 客戶端會在首次連接時完成 OAuth 登入。
{
"mcpServers": {
"china-data-taobao-bilibili-wechat-trendi": {
"type": "http",
"url": "https://mcp.apify.com/?tools=fetch-actor-details,tubetext/taobao-tmall-scraper,tubetext/bilibili-scraper,tubetext/bilibili-video-transcript,tubetext/china-trending-topics,tubetext/wechat-articles-scraper,tubetext/alibaba-yiwu-image-search"
}
}
}README
TubeText MCP servers
Remote MCP servers that give Claude, Cursor, VS Code, ChatGPT and any other MCP client access to TubeText tools: video transcripts, comment sentiment, AI video breakdowns, esports match timelines, ad library data, e-commerce prices, reviews, China data and travel prices.
There is nothing to install or host. Each server is a URL on Apify's hosted MCP server (mcp.apify.com) that exposes a small, focused set of TubeText Actors as tools. You sign in with your own Apify account (OAuth) or pass your Apify API token. Usage is billed under each Actor's current Store pricing and your Apify plan. Review billable events and any optional AI processing before running a tool.
Choose a bundle by task
Use these routes when you know the result you need, rather than an Actor name. Links open the existing bundle definitions; connect with the matching server URL and client instructions.
- Research channels or Shorts: YouTube bundle. Channel Analytics and Social Blade provide creator statistics; Shorts metadata and optional transcripts are separate outputs.
- Find competitor video ads: Ad Intelligence. Meta Ad Library and Google Ads Transparency discover ads; video and shot analysis are separate tools. Ad presence or time running is not evidence of spend, conversions or ROAS.
- Turn episodes or social videos into text: Social Video Transcripts. Choose Podcast Transcript for RSS/Apple Podcasts, or Any Video & Podcast Transcript for its documented mixed-source/direct-media inputs. Public video speech is not tweet, Reddit-comment or private-content scraping; not every URL is supported.
- Research Chinese-language sources: China Data. Bilibili discovery and speech transcription are separate Actors. WeChat covers public Official Account articles, not private chats; trending boards are snapshots, not a historical trend database.
- Build a product or supplier shortlist: E-commerce Prices for marketplace listings; China Data for Alibaba/Yiwu supplier search from a product photo. Visual similarity does not verify the manufacturer or identical goods.
- Analyze customer-review themes: Reviews + AI Insights groups Steam, Google Play and Agoda-displayed hotel reviews. Booking.com-sourced reviews displayed by Agoda do not imply direct Booking.com scraping. AI summaries describe the retrieved sample, not every customer.
- Build a travel or fuel-price dashboard: Travel & Local Prices. Keep dates, currency, fuel grade and source-update timestamps with each comparison. Returned prices are observations, not guaranteed booking or pump prices.
Before calling an Actor, inspect its current input schema and Pricing tab, agree on the input/count and budget, and check Actor Start and optional-processing charges. A bundle definition or directory listing does not verify live tool availability or runtime health.
Looking for Spotify play counts? Use the existing Spotify Actor directly in Apify. It is not included in the MCP bundles listed here.
Servers
Connect
Claude Code
Claude Desktop / claude.ai: Settings → Connectors → Add custom connector → paste a server URL from the list below, then sign in to Apify.
Remote MCP clients (generic example):
Client configuration paths and schemas differ. Use the Apify MCP configurator for your client rather than assuming this JSON works unchanged everywhere. See the VS Code configuration reference for its supported format.
Clients without OAuth: add the header Authorization: Bearer <your Apify API token> (get one here).
Server URLs
- TikTok Transcripts, Comments, Video Breakdown & Shot Analysis:
https://mcp.apify.com/?tools=fetch-actor-details,tubetext/tiktok-transcript,tubetext/youtube-tiktok-comments-api,tubetext/video-breakdown-ai,tubetext/shot-by-shot-video-analysis - YouTube Transcripts, Comments & Channel Analytics:
https://mcp.apify.com/?tools=fetch-actor-details,tubetext/youtube-transcript-fast,tubetext/youtube-shorts-scraper,tubetext/youtube-tiktok-comments-api,tubetext/youtube-channel-analytics,tubetext/socialblade-scraper,tubetext/shot-by-shot-video-analysis - Instagram, Facebook, X, Reddit & Podcast Transcripts:
https://mcp.apify.com/?tools=fetch-actor-details,tubetext/instagram-reels-transcript,tubetext/facebook-video-transcript,tubetext/twitter-x-video-transcript,tubetext/podcast-transcript,tubetext/any-video-podcast-transcript,tubetext/reddit-video-transcript - Meta Ad Library, Google Ads Transparency & Video Ad Analyzer:
https://mcp.apify.com/?tools=fetch-actor-details,tubetext/meta-ad-library-videos,tubetext/google-ads-transparency-video,tubetext/video-ad-analyzer,tubetext/shot-by-shot-video-analysis - E-commerce Prices: Amazon, Taobao, AliExpress, Idealo:
https://mcp.apify.com/?tools=fetch-actor-details,tubetext/amazon-product-search-scraper,tubetext/taobao-tmall-scraper,tubetext/aliexpress-search-scraper,tubetext/tiktok-shop-product-scraper,tubetext/idealo-price-scraper - Reviews + AI Insights: Steam, Agoda, Google Play:
https://mcp.apify.com/?tools=fetch-actor-details,tubetext/steam-reviews-scraper,tubetext/agoda-reviews-scraper,tubetext/google-play-reviews-ai - China Data: Taobao, Bilibili, WeChat, Trending:
https://mcp.apify.com/?tools=fetch-actor-details,tubetext/taobao-tmall-scraper,tubetext/bilibili-scraper,tubetext/bilibili-video-transcript,tubetext/china-trending-topics,tubetext/wechat-articles-scraper,tubetext/alibaba-yiwu-image-search - Travel & Local Prices: Flights, Hotels, Gas:
https://mcp.apify.com/?tools=fetch-actor-details,tubetext/google-flights-scraper,tubetext/google-hotels-scraper,tubetext/gas-prices-scraper - Esports Match Timelines: Valorant & CS2:
https://mcp.apify.com/?tools=fetch-actor-details,tubetext/esports-match-timeline,tubetext/youtube-transcript-fast,tubetext/youtube-tiktok-comments-api
Start with a bounded YouTube use case
Before any paid test, list the available tools and inspect the current input schema for tubetext/youtube-transcript-fast. Connecting or seeing a directory listing does not establish that an Actor run will succeed. If a requested tool is absent, stop and check the selected bundle instead of silently substituting another Actor.
For a first captions-only task, ask:
Get the existing captions for this one public YouTube video: . Set
aiFallbackto false; do not enable AI transcription or run other Actors. Return the transcript with source timestamps, or explain the unsuccessful status without inventing missing text. Show the planned input and cost limit before starting.
AI fallback is enabled by default in the YouTube Actor, so set aiFallback: false explicitly for captions-only work. This does not make the run free. After retrieval, retain the run ID, dataset ID and original export before using the text downstream. For a channel-to-RAG workflow, use the replay-safe caption inbox example, which includes synthetic offline tests rather than a live-service benchmark.
For export steps, read the YouTube transcript export guide: preserve the complete JSON export, save the returned srt field as a subtitle file, and handle WebVTT as a separate conversion.
For the workflow overview and recovery checklist, read the YouTube transcript RAG guide.
Example prompts
- "Get the transcript of this TikTok and break down its hook and shot list: "
- "Break down this ad shot by shot and show where viewers replayed or commented: "
- "Transcribe the last 10 videos on this YouTube channel and summarize the main topics."
- "Which moments of this YouTube video did viewers react to, and how did the top comments feel about them? "
- "Give me the round-by-round score timeline of this Valorant match VOD: "
- "Pull Nike's active video ads from Meta Ad Library and tell me which angles they test."
- "List Duolingo's YouTube video ads in the US from Google Ads Transparency and how long each has run."
- "Get the last 30 Shorts of @MrBeast with views and what is said in the first 3 seconds."
- "Compare prices for this product on Idealo in Germany, France and Italy."
- "What do Steam players complain about most in the recent reviews of ?"
Costs and safety
- Actor calls can incur charges. Check the selected Actor's live Pricing tab: skipping a transcript or analysis event does not guarantee a zero-cost run. Actor Start, data-fetch events, optional AI work or other charges may still apply.
- Agree on a budget before starting, not just for large jobs. Inspect the current tool schema for supported run limits. For Apify's generic
call-actortool, limits such asmaxTotalChargeUsdbelong incallOptions, not in the Actor input (official reference). Do not assume every Actor wrapper exposes the same fields or that a run limit caps downstream model, storage or integration costs. - Apify authentication is required for ordinary account-billed runs. These public-data workflows do not require you to provide a social-network login or cookies. Never paste API tokens into prompts, public issues or committed files.
Feedback
Problem or missing feature? Open an issue on the Actor's Store page (Issues tab) or here. Business or custom work: email us.
Published to the official MCP Registry under io.github.LMSlay/* via GitHub Actions OIDC (.github/workflows/publish-mcp-registry.yml).
來源:README.md,提交 9cee606
工具
0版本歷史
4- v1.3.1最新Oct 19, 2026
- v1.3.0Oct 18, 2026
- v1.2.0Oct 18, 2026
- v1.1.0Oct 14, 2026
評分
已滿足適用檢查項的 85%
該評分衡量此收錄條目的完整程度,依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價,也不表示伺服器安全或目前可用:可連通性是帶日期的檢查結果,評分也不包含安裝量或使用者評價。
- 近 30 天內可連通必需25 分
- 已宣告認證方式必需10 分
- 已列出工具15 分
請讓伺服器在無憑證時回應 tools/list,或發布可在沙箱中無憑證啟動的軟體包。
- 描述必需15 分
- 圖示10 分
- 文件15 分
- 授權條款10 分