Skip to content

Perufitlife/multi-scraper-mcp

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Multi-Scraper MCP — 12 scraping tools for AI agents, one MCP server

Give your AI agent real-time web data with a single connection. This Model Context Protocol server exposes 12 production scraping tools — Reddit, Amazon, eBay, Google Maps (search + reviews), Yelp, YouTube, TikTok, Indeed, Trustpilot, website contact finder, and SaaS pricing tracker — to Claude, ChatGPT, Cursor, Cline, Windsurf, or any MCP client.

One endpoint. Twelve tools. Pay only for the calls your agent makes.

The 12 tools

Tool What your agent gets
reddit_scrape Subreddit posts, Reddit-wide search, comment threads — sentiment & market research
amazon_search Products, prices, ratings, ASINs — 10 marketplaces (US/UK/DE/FR/IT/ES/CA/AU/JP/IN)
ebay_search Live listings + sold listings for real market-price history
google_maps_search Local businesses: name, address, phone, website, rating, GPS
google_maps_reviews Full review text + ratings for any Google Maps place
yelp_search Yelp businesses via the official Fusion API (bring your free Yelp key)
youtube_channel Channel stats + recent video metrics
tiktok_profile Creator profiles, video stats, hashtag discovery
indeed_jobs Job listings with salary data — 7 countries, date/type filters
trustpilot_reviews Company reviews — mine complaints with filterByRating: 1
website_contacts Emails, phones, social links, tech stack from any domain
saas_pricing Structured pricing tiers from any SaaS pricing page + change detection

Quick start

You need an Apify account (free tier works) and your API token from Console → Settings → API & Integrations.

The MCP endpoint (Streamable HTTP):

https://renzomacar--multi-scraper-mcp.apify.actor/mcp

Authenticate with Authorization: Bearer <YOUR_APIFY_TOKEN> (or append ?token=<YOUR_APIFY_TOKEN>).

Claude Code

claude mcp add --transport http multi-scraper \
  "https://renzomacar--multi-scraper-mcp.apify.actor/mcp" \
  --header "Authorization: Bearer YOUR_APIFY_TOKEN"

Claude Desktop

Settings → Developer → Edit Config, then add:

{
  "mcpServers": {
    "multi-scraper": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://renzomacar--multi-scraper-mcp.apify.actor/mcp",
        "--header",
        "Authorization: Bearer YOUR_APIFY_TOKEN"
      ]
    }
  }
}

Cursor / Windsurf / Cline

Add to your MCP config (e.g. .cursor/mcp.json):

{
  "mcpServers": {
    "multi-scraper": {
      "url": "https://renzomacar--multi-scraper-mcp.apify.actor/mcp",
      "headers": { "Authorization": "Bearer YOUR_APIFY_TOKEN" }
    }
  }
}

ChatGPT (Developer Mode connectors)

Settings → Connectors → Advanced → Developer mode → Add custom connector, with URL:

https://renzomacar--multi-scraper-mcp.apify.actor/mcp?token=YOUR_APIFY_TOKEN

Legacy SSE clients

An SSE transport is also available at /sse (+ /messages) for older clients.

Pricing — transparent pay-per-event

Event Price
Tool call (successful) $0.05
Actor start (standby wake-up) $0.00005 per GB
  • No subscription, no minimums — you pay only when your agent actually calls a tool.
  • Underlying scraper runs execute under your Apify account at their own listed rates (typically fractions of a cent per result).
  • Failed tool calls are not charged.

Notes per tool

  • yelp_search needs a Yelp Fusion API key — free, 5,000 calls/day, 2 minutes to get at docs.developer.yelp.com. Pass it as the yelpApiKey tool argument.
  • reddit_scrape survived Reddit shutting down its public .json endpoints (June 2026, universal 403): it now falls back to Reddit RSS feeds with a circuit breaker, so it keeps returning posts. In RSS mode items don't include score/comment counts (source: rss-fallback).
  • Keep maxResults* small (10–30) for snappy agent loops; raise them for batch research.

Example agent workflows

  • Local lead-gen pipeline: google_maps_search ("dentists in Miami") → website_contacts (emails + tech stack) → your CRM.
  • Product intelligence: amazon_search + ebay_search (soldOnly) → real street price vs. listed price.
  • Brand/sentiment monitor: reddit_scrape + trustpilot_reviews (filterByRating: 1) + google_maps_reviews → weekly complaint digest.
  • Competitor watch: saas_pricing on competitor pricing pages with compareWithPrevious: true → alert on changes.
  • Hiring-signal prospecting: indeed_jobs ("Shopify developer") → companies investing in e-commerce → website_contacts.

Why this server

  • Every tool wraps a battle-tested public Apify Actor (4,000+ combined runs).
  • Structured JSON with consistent schemas — built for LLM consumption, results truncated to stay token-friendly.
  • Both modern Streamable HTTP (/mcp) and legacy SSE (/sse) transports.
  • Open source: github.com/Perufitlife/multi-scraper-mcp.

Found this useful?

Please leave a quick review on the Reviews tab — it genuinely helps independent developers get visibility on the Apify Store. Thanks!

About

12 web scraping tools as MCP server for AI agents (Claude Desktop, ChatGPT, Cursor). Reddit, Amazon, eBay, Google Maps, Yelp, YouTube, TikTok, Indeed, Trustpilot, contact finder, SaaS pricing.

Topics

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors