How to give Claude, Cursor or any AI agent web data tools through Apify's MCP server
Short answer
Apify hosts an MCP server at mcp.apify.com. Add it to your MCP client with your Apify token and a ?tools= list of Actors; each Actor becomes one tool with its input schema as arguments. The agent calls it, the run happens on your Apify account, and you pay only per result.
- Get your Apify API token (Apify console, Settings, API & Integrations).
- Choose the Actors the agent may use and join them with commas in ?tools=.
- Add the server to your client config (example below) and restart the client.
- Ask in plain words, e.g. "find the websites of these 20 companies, then their contact emails". The agent chains the tools itself.
- Keep the tool list short: fewer, well-described tools are picked more reliably than dozens.
Copy-paste code
Claude Desktop / Cursor / VS Code: six CyberMax tools in one server
{
"mcpServers": {
"cybermax-web-data": {
"command": "npx",
"args": ["-y", "mcp-remote",
"https://mcp.apify.com/?tools=cybermax/company-to-domain,cybermax/website-contacts,cybermax/domain-intel,cybermax/domain-authority,cybermax/google-trends,cybermax/bulk-image-downloader",
"--header", "Authorization: Bearer ${APIFY_TOKEN}"],
"env": { "APIFY_TOKEN": "<your Apify token>" }
}
}
}Claude Code (one command)
claude mcp add --transport http cybermax-web-data \
"https://mcp.apify.com/?tools=cybermax/company-to-domain,cybermax/website-contacts,cybermax/domain-intel" \
--header "Authorization: Bearer $APIFY_TOKEN"Python agent frameworks: call the same Actor directly
import os
from apify_client import ApifyClient
def company_to_domain(names: list[str]) -> list[dict]:
"""Official website domain for each company name, with a 0-1 confidence."""
c = ApifyClient(os.environ["APIFY_TOKEN"])
run = c.actor("cybermax/company-to-domain").call(run_input={"companies": names})
return list(c.dataset(run["defaultDatasetId"]).iterate_items())
# register company_to_domain as a tool in LangChain, CrewAI, the OpenAI or Anthropic SDK, etc.| Tool | What the agent can do | Price |
|---|---|---|
| cybermax--company-to-domain | Company names to official website domains, with confidence | $0.0008 per match |
| cybermax--website-contacts | A domain's published email, phone, socials and contact page | $0.004 per domain with contacts |
| cybermax--domain-intel | Tech stack, mail provider, DMARC, domain age, TLS expiry | $0.004 per domain |
| cybermax--domain-authority | 0-100 authority score and rank from open web-graph data | $0.002 per ranked domain |
| cybermax--google-trends | Rising/falling verdict, rising and breakout queries | $0.003 per keyword |
| cybermax--bulk-image-downloader | Every image on a list of pages, full size, in a ZIP | $0.0008 per image |
| cybermax--website-to-markdown | A whole website or docs site as clean Markdown, RAG chunks and llms.txt | $0.001 per page |
| cybermax--seo-audit | Every page's SEO issues, broken links and a 0-100 score | $0.004 per page |
| cybermax--shopify-scraper | A Shopify or WooCommerce store's products, prices and stock | $0.0008 per product |
| cybermax--remote-jobs | Fresh remote jobs with salaries from Remote OK and Jobicy | $0.0008 per job |
| cybermax--image-converter | Image URLs to WebP, AVIF, JPEG or PNG, resized, in a ZIP | $0.0015 per image |
Source: tools/list on https://mcp.apify.com/?tools=cybermax/<actor>, 10 Nov 2026; prices from each Actor's Apify Store page
A typical agent chain
Names to domains (company-to-domain), domains to contact pages (website-contacts), and the stack and size of each company (domain-intel, domain-authority): the agent builds a qualified account list from a plain list of names, and every step is a paid-per-result call you can audit in the Apify console.
Let the agent read whole websites
Add cybermax/website-to-markdown (Docsieve) to the tools list and the agent can read a docs site or website as clean Markdown before it answers, or build heading-aware chunks and llms.txt for a RAG index. Pages cost $0.001 each; the llms.txt files are free.
Also in the official MCP Registry
The same Apify-hosted tools are listed in the official MCP Registry under the dev.workers.cybermax-tools.cybermax namespace, so registry-aware clients and directories can find them by name.
Good to know
Runs are billed to your own Apify account at each Actor's published per-result price; Apify's free plan includes $5 of usage each month.
Without the Authorization header the endpoint answers 401 by design.
Namewhere: Company names in, official website domains out
$0.0008 per match, misses freeStart with Namewhere: company names in, official website domains out, the first step of most account-research agents. $0.0008 per match; misses are free.
Runs on your own Apify account, pay per result; failed results are not charged.Questions
Do I need to install anything?
Only an MCP client. mcp-remote (via npx) bridges clients that do not speak remote HTTP MCP yet; clients with native remote MCP support can use the URL and header directly.
Which clients work?
Any MCP client: Claude Desktop, Claude Code, Cursor, VS Code, Windsurf and agent frameworks with MCP support.
Can I limit cost?
Yes: each tool call states its input size, and you can set a maximum spend per run in Apify. Misses are free on the six tools in the config above.
Free: the Free API Starter Kit
Copy-paste calls for 4 APIs with free daily tiers (CVE priority, VIN recalls, US duty rates, email checks), then 5 short emails on KEV/EPSS, MCP and free data.