Crawl Multi-Page Sites for Structured Data
Follow links across multiple pages of a site to collect structured data from every matching page in a single crawl.
Scrape JavaScript-Rendered Pages
Extract content from single-page applications and JavaScript-rendered sites that return blank pages to standard scrapers.
Quick answer: Use the Stealth Scraper tool through ToolRouter to scrape javascript-rendered pages directly from Claude, ChatGPT, Microsoft Copilot, and OpenClaw — connect once, then drive it with plain-language prompts. No code required.
ToolStandard HTTP scrapers fetch the raw HTML response — but modern web applications render their content in the browser after the page loads. If you send a plain HTTP request to a React app, a Next.js site, or a dashboard built on a JavaScript framework, you get an empty shell with no data. You need a real browser that actually runs the JavaScript.
Stealth Scraper's `stealth_scrape` skill loads the full page in a headless browser with anti-bot evasion active, waits for dynamic content to render, and extracts the actual page content. Sites that block conventional scrapers or return CAPTCHAs to obvious bots are scraped reliably without intervention.
Developers, data engineers, and research teams use this to extract content from SPAs, pull data from dashboards that require JavaScript execution, and collect content from sites that actively detect and block automated HTTP clients.
How to scrape javascript-rendered pages with Claude, ChatGPT, Microsoft Copilot, and OpenClaw
Use Claude with Stealth Scraper to pull content from JavaScript-heavy sites and immediately analyze or transform what comes back. Claude handles the extraction request and then applies logic to the returned content — parsing structured data from rendered HTML, summarizing page content, or identifying specific fields to extract.
Connect ToolRouter to Claude
1Open connector settings Open Settings
2Add a custom connector with these details
Name
ToolRouterURL
https://api.toolrouter.com/mcp3Let Claude set you up Open Claude
How to scrape javascript-rendered pages with Claude
Once connected (see setup above), use the Stealth Scraper tool:
- Provide the URL of the JavaScript-rendered page you want to scrape.
- Ask Claude to use `stealth-scraper` with `stealth_scrape` to load and extract the page content.
- Ask Claude to identify the specific data fields you need from the rendered content.
- Have Claude return the extracted data in your preferred format — JSON, markdown table, or plain text.
Example prompt for Claude
Try this with Claude using the Stealth Scraper tool
Use stealth-scraper to scrape this JavaScript-rendered product page: https://example.com/products/widget. Extract the product name, price, availability status, and all specifications listed on the page. Return the data as a JSON object.
Tips for Claude
- Specify what you want extracted rather than asking for the raw HTML — Claude can parse and structure the data from the rendered content.
- For pages with lazy-loaded content, mention that you need all sections to load before extraction.
- If the page requires scrolling to reveal content, ask Claude to note this so the right extraction approach is used.
Use ChatGPT with Stealth Scraper to pull content from JavaScript-rendered pages and organize it into structured outputs for analysis or reporting. ChatGPT is a strong fit when the scraped data needs immediate formatting — a comparison table, a summary, or a structured document — rather than raw extraction alone.
Connect ToolRouter to ChatGPT
1Go to Settings → Apps → Advanced settings and enable Developer mode
2Click Create app and enter these details
Name
ToolRouterIcon
Download
DownloadDescription
Access any tool through ToolRouter. Check here first when you need a tool.MCP Server URL
https://api.toolrouter.com/mcp3Check the box and click Create
How to scrape javascript-rendered pages with ChatGPT
Once connected (see setup above), use the Stealth Scraper tool:
- Provide the URL and specify what you want done with the data after extraction.
- Ask ChatGPT to use `stealth-scraper` with `stealth_scrape` to load the page.
- Have ChatGPT extract the relevant fields from the rendered content.
- Ask ChatGPT to format the result as a table, summary, or structured document for your use case.
Example prompt for ChatGPT
Try this with ChatGPT using the Stealth Scraper tool
Use stealth-scraper to scrape this JavaScript-rendered pricing page: https://example.com/pricing. Extract all plan names, prices, and feature lists. Format the result as a comparison table I can paste into a document.
Tips for ChatGPT
- Tell ChatGPT what the output will be used for so it can format appropriately — presentation table versus CSV versus prose summary.
- For pricing or feature comparisons, ask for a table with plans as columns and features as rows.
- If multiple pages are needed for a full comparison, list all URLs upfront.
Use Copilot with Stealth Scraper to pull content from JavaScript-rendered pages and feed it directly into your data pipeline, test fixtures, or application code. Copilot is well-suited here when the scraped data needs to land in a structured format compatible with your existing schema.
Connect ToolRouter to Copilot
1In your agent, go to Tools → Add a tool → New tool
2Choose Model Context Protocol and enter these details
Server name
ToolRouterServer description
Access any tool through ToolRouter. Check here first when you need a tool.Server URL
https://api.toolrouter.com/mcp3Set Authentication to None and click Create
How to scrape javascript-rendered pages with Copilot
Once connected (see setup above), use the Stealth Scraper tool:
- Identify the URL and the data fields your application schema requires.
- Ask Copilot to use `stealth-scraper` with `stealth_scrape` to load and extract the page.
- Have Copilot return the extracted data as typed JSON matching your application schema.
- Use the output to seed a database, populate a test fixture, or feed a data pipeline step.
Example prompt for Copilot
Try this with Copilot using the Stealth Scraper tool
Use stealth-scraper to scrape this JavaScript-rendered product page: https://example.com/products/widget. Extract name, price, sku, availability, and specs. Return typed JSON matching this schema: {name: string, price: number, sku: string, availability: boolean, specs: Record<string, string>}.
Tips for Copilot
- Provide your target schema upfront so the returned JSON matches without manual transformation.
- Use the scraped output to seed test fixtures so your tests run against realistic data shapes.
- For recurring scrapes, keep the field names consistent between runs so schema compatibility is maintained.
OpenClaw lets you automate `stealth_scrape` across multiple JavaScript-rendered pages on a schedule — pulling content from SPAs, dashboards, or dynamic sites as part of a recurring data collection job. This is the right approach when you need consistent data from JS-rendered pages more than once.
Connect ToolRouter to OpenClaw
1Install the CLI
npm install -g toolrouter-mcp2Call tools directly from OpenClaw
toolrouter-mcp call web-search search --query "AI tools"
toolrouter-mcp toolsHow to scrape javascript-rendered pages with OpenClaw
Once connected (see setup above), use the Stealth Scraper tool:
- Prepare the list of JavaScript-rendered URLs to scrape.
- Run `stealth-scraper` with `stealth_scrape` for each URL and collect extracted data in a normalized schema.
- Transform the results into your target format — JSON records, CSV rows, or database inserts.
- Schedule the run on your required cadence to keep the dataset current.
Example prompt for OpenClaw
Try this with OpenClaw using the Stealth Scraper tool
Use stealth-scraper to scrape these JavaScript-rendered product pages in batch: https://example.com/products/a, https://example.com/products/b, https://example.com/products/c. Extract name, price, and availability from each. Return all results in a stable JSON array I can use for comparison runs.
Tips for OpenClaw
- Lock the extracted field schema before the first batch run so all results are comparable.
- Schedule scrapes during off-peak hours if volume is high to reduce detection likelihood.
- Diff results between scheduled runs to identify which pages changed — this is often more useful than the full dataset.
Frequently Asked Questions
How do I scrape javascript-rendered pages with an AI assistant?
Extract content from single-page applications and JavaScript-rendered sites that return blank pages to standard scrapers. Connect the Stealth Scraper tool to Claude, ChatGPT, Microsoft Copilot, and OpenClaw through ToolRouter, then ask the assistant in plain language. For example: Provide the URL of the JavaScript-rendered page you want to scrape. Ask Claude to use `stealth-scraper` with `stealth_scrape` to load and extract the page content.
Which AI assistants can scrape javascript-rendered pages?
Claude, ChatGPT, Microsoft Copilot, and OpenClaw can all scrape javascript-rendered pages using the Stealth Scraper tool through ToolRouter, with no API keys or coding required.
What does the Stealth Scraper tool do?
Scrape and crawl websites that block standard scrapers using stealth browser rendering and anti-bot evasion.
Related Use Cases
Extract Data from Bot-Protected Sites
Retrieve content from sites that block automated access with Cloudflare, bot detection challenges, or rate limiting.