Website Phrase and Business Model Classifier MCP Server
Classify company websites from their public pages using deterministic rules via Apify.\n\n- Use the classify_website tool with one domain or up to 500 domains.\n- Optionally provide up to 25 phrases to detect buyer phrases, with matched and negated phrase evidence.\n- Read the homepage plus up to 5 linked pages by default, configurable via max_pages from 1 to 10.\n- Return business model classification: b2b, b2c, both, or unknown.\n- Return category flags: SaaS, ecommerce, agency, manufacturer, wholesaler, distributor, and marketplace.\n- Report evidence text and URL, plus row_status for unreachable, blocked, robots.txt-disallowed, empty, or error sites.\n- Filter categories, choose proxy_mode (auto or direct), and optionally set max_charge_usd.\n- Requires an APIFY_TOKEN; read-only tool; Apify charges per classified domain.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Website Phrase and Business Model Classifier MCP Server Classify stripe.com as B2B SaaS and shopify.com as B2C store"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Website Phrase and Business Model Classifier MCP Server
An MCP server for the Mamba Labs Website Phrase and Business Model Classifier actor on Apify. Give it company domains. For each one it reads the homepage and up to 5 pages it links to, and returns one row: B2B, B2C, or both; whether the site presents itself as SaaS, an online store, an agency, a manufacturer, a wholesaler, a distributor, or a marketplace; and whether your own phrases appear, with the quoted text and the page as evidence.
What's Inside
Related MCP server: AI Tooling Detector MCP Server
What it does
The rules are fixed and written out, not a language model, so the same pages always give the same answer. A site that does not say enough comes back unknown. A hit inside a negation ("we are not an agency") never counts. Unreachable, blocked, robots.txt disallowed, and empty sites come back free with a row_status saying so. All of the work runs on Apify. This package is a thin client that starts the actor run, waits for it, and hands back the dataset rows.
Quick start
You need Node.js 18 or newer and an Apify account with an API token.
Add this to your Claude Desktop config (claude_desktop_config.json), Cursor, or any MCP client:
{
"mcpServers": {
"mamba-website-phrase-business-model-classifier": {
"command": "npx",
"args": [
"-y",
"@mambalabsdev/mcp-website-phrase-business-model-classifier"
],
"env": {
"APIFY_TOKEN": "your-apify-token"
}
}
}
}Restart the client. The classify_website tool appears.
Prerequisites
Node.js 18 or newer.
An Apify API token from console.apify.com/account/integrations, set as
APIFY_TOKEN. The server lists its tool without a token; a call needs one.
Example prompts
"Which of these 40 domains are agencies, and which of them say they are a Clay agency?"
"Split this store list into B2B and B2C and show me the evidence for each."
"Do any of these suppliers say they do private label?"
Tool
classify_website. One tool. Example arguments:
{
"domains": ["lodgecastiron.com", "hubspot.com"],
"phrases": ["cast iron", "private label"]
}Inputs
Every buyer facing input of the actor, unchanged. Omitted inputs take the actor's own defaults. Numbers are sent as strings.
Input | Type | Description |
| array | Company domains, up to 500 |
| string | One domain |
| array | Your buyer phrases, up to 25 |
| array | Any of saas, ecommerce, agency, manufacturer, wholesaler, distributor, marketplace; empty means all |
| string | Pages per site, 1 to 10. Default 6 |
| string | Optional cost cap for the run |
| string | auto (default) or direct |
Pricing
Pay per event on Apify, billed to your Apify account. Prices in USD.
Event | Free plan | Bronze | Silver | Gold |
| $0.0045 | $0.004275 | $0.00405 | $0.003825 |
| $0.0005 | $0.0005 | $0.0005 | $0.0005 |
Unreachable, blocked, robots.txt disallowed, empty, and error rows are free.
Your keys
The actor needs no key and holds none. It reads the company's own public web pages. Proxy traffic adds nothing to your price: Mamba Labs pays the platform costs of the run.
Full actor documentation
apify.com/mambalabs/website-phrase-business-model-classifier
License
MIT. See LICENSE.
Available Tools
1 toolclassify_websiteClassify company websitesARead-only
Classify company websites from their own pages. Give domains (up to 500) and optionally your own phrases. For each domain the actor reads the homepage and up to 5 linked pages (about, products, services, wholesale, manufacturing, industries, pricing) and returns one row: business_model b2b, b2c, both, or unknown with the signals found; is_saas, is_ecommerce, is_agency, is_manufacturer, is_wholesaler, is_distributor, is_marketplace; phrases_matched and phrases_negated; and top_evidence_text with top_evidence_url quoting the page. Rules are deterministic, no language model. Unreachable, blocked, robots.txt disallowed, and empty sites come back free with a row_status saying so. A run that does not succeed is returned as an error with its run ID. Read-only; requires an APIFY_TOKEN; Apify charges per domain classified.
Inputs: domain, domains, phrases, max_pages, +3 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 1 tool
Only one tool exists, so there is no possibility of selecting the wrong tool. Its purpose—classifying websites by business model and phrase matches—is clearly stated.
The single tool follows a clear snake_case verb_noun convention: classify_website. There are no competing naming styles to create inconsistency.
A single batch-oriented classification tool is well matched to this narrow, read-only domain. The count is lower than the typical 3–15 range, but the server's scope does not demand separate operations.
The tool covers the core classification workflow with detailed per-domain outputs and statuses for unreachable or blocked sites. Minor gaps exist, such as no dedicated run-status or results-retrieval tool, though the synchronous response largely obviates that need.
Maintenance
Related MCP Connectors
Batch website technology detection with evidence. Paid Apify Actor; bring your own account.
Find emails, phone numbers and social profiles on a list of company websites. One row per domain.
Turn a company domain into a structured B2B sales brief from their public website.
Tech stack, screenshots, WHOIS, SEO audit, sitemaps and AI readiness for any site, via Apify.
Related MCP Servers
- AlicenseAqualityDmaintenanceEnables scraping of emails, phone numbers, and social profile links from website domains. Supports batch processing of up to 20 domains and can find company websites by keyword/company name.2MIT
- AlicenseAqualityBmaintenanceEnables AI clients to analyze any company domain and determine its AI maturity tier (commercialized, deployed, declared, or none) with supporting evidence, via a single tool backed by an Apify actor.173 npmMIT
- AlicenseAqualityBmaintenanceEnables users to convert a vendor partner directory (such as PartnerPage or the Shopify Partner Directory) or a trade body member list into one company row per listing, each carrying website, domain, tier, tags, and location for Clay-ready output. It runs the extraction on Apify, returning dataset rows for up to 20 directory URLs with optional deduping, detail enrichment, and a cost cap.165 npmMIT
- AlicenseNot gradedqualityCmaintenanceEnables auditing one or many B2B SaaS domains to classify their sales motion and pricing model, extract starting USD prices, free-tier/trial availability, billing stack, and whether SAML/SSO is gated behind an Enterprise tier.MIT