# AGENTS.md: Scrapewise

## What is Scrapewise?
Scrapewise turns websites into clean, matched data: scrape, clean, match, deliver. It has ready data
APIs (35 endpoints for Amazon, Walmart, Google, YouTube and ChatGPT; Google Trends interest by region is coming soon), AI scrapers for any other site,
data cleaning and product matching. Billing is pay as you go from a prepaid wallet.

## Capabilities
- Data APIs: send a URL or a search term, get rows back. Price per call for every endpoint: https://scrapewise.ai/scrapers/data-apis
- AI scrapers for any other site: paste a link and the page becomes rows. Input can be one link, a link list or sitemap, or an uploaded file of URLs (CSV or JSON Lines)
- Scheduled runs (daily, every few days, weekly or monthly) and on-demand runs
- Cleaning with group rules: price per unit (for example per kg or per 100 g) and prices converted to EUR
- Product matching against your own product list: text columns first, then product pictures
- Output: REST API, CSV, Excel, or MCP

## API access
- REST API: live, base URL https://portal.scrapewise.ai/api/scraper-api, docs at https://docs.scrapewise.ai/rest-api/overview
- MCP server: live at https://mcp.scrapewise.ai/mcp (Claude Code, Claude Desktop and other MCP clients)
- Authentication: API key (Bearer token), created in the portal under Settings > API Keys
- MCP keys: a read-only key reads saved rows and job results at no cost. A full-access key can also call the paid tools, run and create scrapers and set group schedules.
- MCP data tools: find_local_businesses (Google Maps places, up to 20 per call, 10 credits per call, no e-mail addresses), google_search (web, news or shopping results, 10 credits per query), search_amazon (1 credit per result page), get_amazon_product (Amazon product pages by ASIN or address, 1 credit per product, plus 1 credit per product for its offer listing on request), amazon_seller_leads (one row per Amazon seller for a search phrase, with the business details the seller published: company, email, phone, address and VAT number; sellers on amazon.de almost always publish an email, sellers on amazon.co.uk and amazon.com do not; 1 credit per page read, about 3 credits per seller, always a job), idealo_offers (every shop's price, total price, rating and offer link for a product on idealo, by EAN or idealo product page, markets de, at, fr, it, es and uk, 10 credits per product), google_trends (interest over time, 10 credits per call), fetch_page (text or HTML of one page, usually 5 credits, at most 25), get_product_data (product data from the product pages of one shop, usually 5 credits per page, at most 25), clean_product_data (price in EUR, price per unit and duplicate removal on rows you already have, free, fetches nothing), get_results and cancel_job for long jobs (free), scrapewise_get_scraper_group_list and scrapewise_get_scraper_data_group_client to read saved rows (free). 1 credit = €0.00015.
- MCP price first: every paid tool takes quote_only=true, which returns the price and the wallet balance without spending anything. That price is the most the call can cost, and the result of every paid call reports its cost. The paid tools and clean_product_data take their arguments inside a body object. Send the same idempotency_key on a retry to be charged once. Long requests return a job_id for get_results.
- MCP keys: A Full key (LLM_FULL) lists every tool. A Read key (LLM_READ) lists the read tools.
- Rate limits: per API key. Over the limit the API returns 429 with a Retry-After header.
- MCP setup page, with one-click install for Claude Code, Cursor and VS Code: https://scrapewise.ai/mcp
- MCP registry manifest (server.json): https://scrapewise.ai/server.json
- MCP namespace proof: https://scrapewise.ai/.well-known/mcp-registry-auth
- Agent skills: https://scrapewise.ai/skills

## Pricing (machine-readable)
See: https://scrapewise.ai/pricing.md
Short version: prepaid wallet in euros, 5 free requests on every new account, top-ups of €10, €25 or €50
(or any amount from €5 to €1,000), from €0.15 per 1,000 plain pages. No plan, no subscription.

## Use cases
- Competitor price tracking: https://scrapewise.ai/use-cases/competitor-price-tracking
- Product matching: https://scrapewise.ai/use-cases/product-matching
- MAP monitoring and unauthorized sellers: https://scrapewise.ai/use-cases/map-monitoring
- Supplier price monitoring: https://scrapewise.ai/use-cases/supplier-price-monitoring
- Product catalogue extraction: https://scrapewise.ai/use-cases/product-data-extraction
- Data for AI and LLMs: https://scrapewise.ai/use-cases/extract-data-for-ai-and-llms

## Example
bebo.ee, a grocery price comparison across 12 e-shops, runs on Scrapewise: shop prices are scraped, cleaned to
price per kg and matched across chains.

## Contact
- Website: https://scrapewise.ai
- Support: siim.brazier@scrapewise.ai
- Founder: Siim Brazier (siim@scrapewise.ai)
