菜单
Search the web, scrape websites, extract structured data from URLs, and automate browsers using Bright Data's Web MCP. Use when fetching live web content, bypassing blocks/CAPTCHAs, getting product data from Amazon/eBay, social media posts, or when standard requests fail.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/patchy631/ai-engineering-hub --skill brightdata-web-mcp命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
Fine-tune a model with GRPO on Fireworks-managed GPUs from a plain-English task description and a dataset. Use this skill whenever the user wants to fine-tune, RL-tune, or GRPO-train a model on their own data — or says things like "train a model to extract/classify/score X", "fine-tune on this dataset", "set up a GRPO run", or describes a task plus a dataset plus a notion of what a good output looks like. Trigger even when the user does not name GRPO or Fireworks explicitly.
This skill should be used when users want to run any workload on Hugging Face Jobs infrastructure. Covers UV scripts, Docker-based jobs, hardware selection, cost estimation, authentication with tokens, secrets management, timeout configuration, and result persistence. Designed for general-purpose compute workloads including data processing, inference, experiments, batch jobs, and any Python-based tasks. Should be invoked for tasks involving cloud compute, GPU workloads, or when users mention running jobs on Hugging Face infrastructure without local setup.
This skill should be used when users want to train or fine-tune language models using TRL (Transformer Reinforcement Learning) on Hugging Face Jobs infrastructure. Covers SFT, DPO, GRPO and reward modeling training methods, plus GGUF conversion for local deployment. Includes guidance on the TRL Jobs package, UV scripts with PEP 723 format, dataset preparation and validation, hardware selection, cost estimation, Trackio monitoring, Hub authentication, and model persistence. Should be invoked for tasks involving cloud GPU training, GGUF conversion, or when users mention training on Hugging Face Jobs without local GPU setup.
基于 SOC 职业分类
| name | brightdata-web-mcp |
| description | Search the web, scrape websites, extract structured data from URLs, and automate browsers using Bright Data's Web MCP. Use when fetching live web content, bypassing blocks/CAPTCHAs, getting product data from Amazon/eBay, social media posts, or when standard requests fail. |
Use this skill for reliable web access in MCP-compatible agents. Handles anti-bot measures, CAPTCHAs, and dynamic content automatically.
Tool: search_engine
Input: { "query": "latest AI news", "engine": "google" }
Returns JSON for Google, Markdown for Bing/Yandex. Use cursor parameter for pagination.
Tool: scrape_as_markdown
Input: { "url": "https://example.com/article" }
Tool: extract
Input: {
"url": "https://example.com/product",
"prompt": "Extract: name, price, description, availability"
}
| Scenario | Tool | Mode |
|---|---|---|
| Web search results | search_engine | Rapid (Free) |
| Clean page content | scrape_as_markdown | Rapid (Free) |
| Parallel searches (up to 10) | search_engine_batch | Pro/advanced_scraping |
| Multiple URLs at once | scrape_batch | Pro/advanced_scraping |
| HTML structure needed | scrape_as_html | Pro/advanced_scraping |
| AI JSON extraction | extract | Pro/advanced_scraping |
| Dynamic/JS-heavy sites | scraping_browser_* | Pro/browser |
| Amazon/LinkedIn/social data | web_data_* | Pro |
Remote (recommended) - No installation required:
SSE Endpoint:
https://mcp.brightdata.com/sse?token=YOUR_API_TOKEN
Streamable HTTP Endpoint:
https://mcp.brightdata.com/mcp?token=YOUR_API_TOKEN
Local:
API_TOKEN=<token> npx @brightdata/mcp
search_engine, scrape_as_markdown&pro=1 to URLPRO_MODE=trueSelect specific tool bundles instead of all Pro tools:
&groups=ecommerce,socialGROUPS=ecommerce,social| Group | Description | Featured Tools |
|---|---|---|
ecommerce | Retail & marketplace data | web_data_amazon_product, web_data_walmart_product |
social | Social media insights | web_data_linkedin_posts, web_data_instagram_profiles |
browser | Browser automation | scraping_browser_* |
business | Company intelligence | web_data_crunchbase_company, web_data_zoominfo_company_profile |
finance | Financial data | web_data_yahoo_finance_business |
research | News & dev data | web_data_github_repository_file, web_data_reuter_news |
app_stores | App store data | web_data_google_play_store, web_data_apple_app_store |
travel | Travel information | web_data_booking_hotel_listings |
advanced_scraping | Batch & AI extraction | scrape_batch, extract, search_engine_batch |
Cherry-pick individual tools:
&tools=scrape_as_markdown,web_data_linkedin_person_profileTOOLS=scrape_as_markdown,web_data_linkedin_person_profileNote:
GROUPSorTOOLSoverridePRO_MODEwhen specified.
search_engine - Google/Bing/Yandex SERP results (JSON for Google, Markdown for others)scrape_as_markdown - Clean Markdown from any URL with anti-bot bypasssearch_engine_batch - Up to 10 parallel searchesscrape_batch - Up to 10 URLs in one requestscrape_as_html - Full HTML responseextract - AI-powered JSON extraction with custom promptsession_stats - Monitor tool usage during sessionFor JavaScript-rendered content or user interactions:
| Tool | Description |
|---|---|
scraping_browser_navigate | Open URL in browser session |
scraping_browser_go_back | Navigate back |
scraping_browser_go_forward | Navigate forward |
scraping_browser_snapshot | Get ARIA snapshot with element refs |
scraping_browser_click_ref | Click element by ref |
scraping_browser_type_ref | Type into input (optional submit) |
scraping_browser_screenshot | Capture page image |
scraping_browser_wait_for_ref | Wait for element visibility |
scraping_browser_scroll | Scroll to bottom |
scraping_browser_scroll_to_ref | Scroll element into view |
scraping_browser_get_text | Get page text content |
scraping_browser_get_html | Get full HTML |
scraping_browser_network_requests | List network requests |
Pre-built extractors for popular platforms:
E-commerce:
web_data_amazon_product, web_data_amazon_product_reviews, web_data_amazon_product_searchweb_data_walmart_product, web_data_walmart_sellerweb_data_ebay_product, web_data_google_shoppingweb_data_homedepot_products, web_data_bestbuy_products, web_data_etsy_products, web_data_zara_productsSocial Media:
web_data_linkedin_person_profile, web_data_linkedin_company_profile, web_data_linkedin_job_listings, web_data_linkedin_posts, web_data_linkedin_people_searchweb_data_instagram_profiles, web_data_instagram_posts, web_data_instagram_reels, web_data_instagram_commentsweb_data_facebook_posts, web_data_facebook_marketplace_listings, web_data_facebook_company_reviews, web_data_facebook_eventsweb_data_tiktok_profiles, web_data_tiktok_posts, web_data_tiktok_shop, web_data_tiktok_commentsweb_data_x_postsweb_data_youtube_videos, web_data_youtube_profiles, web_data_youtube_commentsweb_data_reddit_postsBusiness & Finance:
web_data_google_maps_reviews, web_data_crunchbase_company, web_data_zoominfo_company_profileweb_data_zillow_properties_listing, web_data_yahoo_finance_businessOther:
web_data_github_repository_file, web_data_reuter_newsweb_data_google_play_store, web_data_apple_app_storeweb_data_booking_hotel_listingssearch_engine to find relevant URLsscrape_as_markdown to get contentextract for structured JSON (if needed)web_data_amazon_product for structured product dataweb_data_amazon_product_reviews for review analysisweb_data_* tools for structured extractionscrape_as_markdown + extractscraping_browser_navigate → open URLscraping_browser_snapshot → get element refsscraping_browser_click_ref / scraping_browser_type_ref → interactscraping_browser_screenshot → capture results| Variable | Description | Default |
|---|---|---|
API_TOKEN | Bright Data API token (required) | - |
PRO_MODE | Enable all Pro tools | false |
GROUPS | Comma-separated tool groups | - |
TOOLS | Comma-separated individual tools | - |
RATE_LIMIT | Request rate limit | 100/1h |
WEB_UNLOCKER_ZONE | Custom zone for scraping | mcp_unlocker |
BROWSER_ZONE | Custom zone for browser | mcp_browser |
web_data_* tools when available (faster, more reliable)scrape_as_markdown + extract for unsupported sitesscrape_batch, search_engine_batch)session_statsUse full Node.js path instead of npx:
"command": "/usr/local/bin/node",
"args": ["node_modules/@brightdata/mcp/index.js"]
web_data_* tools (often faster)For detailed documentation, see: