FirecrawlreadA Hand can already work with text you give it. Firecrawl removes a common step before that work starts: manually opening pages, copying useful information and explaining how a public website is structured.
With Firecrawl connected, your Hand can search the public web, read a public page, discover the pages that make up a site and — when you approve it — run a small bounded crawl across part of that site.
Firecrawl is most useful when the public web contains information you need for a new piece of work. Your Hand can collect it, understand it and turn it into something appropriate for your own website, audience and brand.
What Firecrawl adds to a Hand
- Search: find useful public pages.
- Read: turn one public page into clean text and links the Hand can work with.
- Map: see which pages exist on a public site before deciding what matters.
- Crawl: with review, read a bounded group of pages when one page is not enough.
The important part is what happens next. Your Hand can compare the information, organize it, combine it with your own files and turn it into a concrete result.
Use case 1: move an old website without starting from zero
An old website may have outdated design or technology while still containing years of useful service pages, FAQs and explanations. Firecrawl can map the public site and read the pages worth keeping as input for a cleaner replacement.
“Map our current public website. Find the service pages, FAQs and company information worth keeping. Propose a simpler structure for the new site and rewrite the useful material for it. Ignore old news and duplicate pages.”
If the replacement site lives in GitHub, the Hand can then turn that plan into the new pages. Firecrawl supplies the source material; GitHub becomes the new source of truth.
Use case 2: turn manufacturer information into content that sounds like you
A manufacturer may have the specifications and technical explanation for a product, but their page is rarely written for your exact customer or in your tone of voice.
Your Hand can read the public product page, keep the useful facts accurate and create fresh content around them: a clearer description, buying advice, a comparison or an FAQ for your own audience.
“Read the manufacturer's page for this product. Keep the specifications accurate, but write a new product page in our style for homeowners who do not know the technical jargon. Add five practical questions customers are likely to ask.”
Public information is still somebody else's content. Use information you are entitled to use and ask your Hand to create new wording rather than reproducing the source page.
More useful ways to use Firecrawl
- Clean up a product catalogue. Read supplier pages and normalize product names, specifications and descriptions into one consistent structure.
- Research a market. Find public competitor and industry pages, then summarize recurring customer questions, claims and positioning.
- Refresh your own website. Map the site, spot overlapping or outdated pages and decide what should be rewritten, merged or removed.
- Build better FAQs. Read public manuals, help pages or documentation and turn the useful details into clear customer-facing answers.
- Prepare comparisons. Gather public product or service information from several sources and organize the differences into one useful overview.
- Understand public documentation faster. Let the Hand read the relevant documentation before it writes instructions, support material or implementation notes for you.
Tell the Hand what the source is for
“Scrape this website” is technically clear, but not very useful by itself. Give the Hand the outcome you want. Then it can decide whether it needs one page, a site map or a small crawl.
“We are replacing this old website. Find the pages with useful customer information, ignore news and boilerplate, and propose what should survive in the new structure.”
“Research how five competitors explain this service. Do not copy their wording. Show me the common questions they answer and which useful customer questions seem to be missing.”
Know the limits
Firecrawl reads the public web. It is not an export button for an entire business system.
- No private areas. It does not recover logged-in content, a CMS database or a customer portal.
- No backend migration. Forms, checkout logic, accounts, databases and integrations still need to be rebuilt deliberately.
- Crawls are bounded. Hands limits a reviewed Firecrawl crawl to at most 20 pages and depth 2. Crawls consume Firecrawl credits.
- Firecrawl is currently a preview connection. Availability can differ by Hands account.
Firecrawl is most valuable before the writing starts.
The connector is a bridge between information that already exists on the public web and the work you actually want from your Hand. That can be a new website, better product content, a comparison, a research brief, a cleaner FAQ or simply a faster way to understand a site you do not want to read page by page yourself.
Give your Hand public source material it can actually work with.
Connect Firecrawl when you want the Hand to research, map or read public websites as part of a larger task. The connection is currently available as a bounded preview.