Business Data MCP Server

Gov contracts, jobs, LinkedIn and company data as tools for AI agents. Hosted on Apify: apify.com/artificially/business-data-mcp

One MCP server that gives Claude, ChatGPT, Cursor and any other AI agent 13 business data tools: US federal contract opportunities and awards, EU public tenders, jobs from Google Jobs, LinkedIn and company career sites, LinkedIn company and person profiles, Y Combinator startups, company contact details, website tech stacks, PDF text extraction and Google Trends.

Each tool runs one of our dedicated Artificially Actors on your Apify account and returns compact JSON sized for an agent's context window.

Built by Artificially | Follow us on X @artificially_io

Why this MCP server

• One connection, 13 tools – no need to add each Actor separately. The agent sees clear tool names, short descriptions and small, typed inputs.
• Agent-friendly output – every tool asks the underlying Actor for its compact output, keeps result counts small by default and caps the response size, so answers fit in the context window.
• Handles slow runs – long runs report progress; if a run is still going when the wait limit is reached, the tool returns what is ready plus a runId, and the agent calls get_run_results later.
• Clear errors – failed runs, missing credit, blocked domains and unreadable documents come back as short messages the agent can act on.
• You only pay for results – no extra fee for the MCP layer (see Pricing).

Tools

Tool	What it does	Backed by
search_gov_contract_opportunities	Open US federal solicitations, RFPs and sources sought on SAM.gov by keyword, NAICS, agency, set-aside and date	SAM.gov Opportunities Scraper
search_awarded_federal_contracts	Awarded US federal contracts (winners, amounts, agencies) from USAspending.gov	US Federal Contracts Scraper
search_eu_tenders	EU/EEA public tenders and contract awards from TED by keyword, country, CPV and notice type	EU Tenders Scraper
search_company_career_jobs	Open roles from company career sites (Greenhouse, Lever, Ashby, Workable and more)	Career Site Jobs Scraper
search_google_jobs	Google Jobs search results aggregated from job boards and company sites	Google Jobs Scraper
search_linkedin_jobs	Public LinkedIn job postings by keywords, location, date, workplace type and seniority	LinkedIn Jobs Scraper
get_linkedin_company	LinkedIn company pages: website, industry, size, employees, followers, HQ, funding	LinkedIn Company Scraper
get_linkedin_profiles	Public LinkedIn person profiles: headline, current company, experience, education	LinkedIn Profile Scraper
find_yc_companies	Y Combinator startups by keyword, batch, industry, status and hiring, optionally with founders	Y Combinator Scraper
get_company_contacts	Emails, phone numbers, social profiles and contact pages from company websites	Company Contact Extractor
detect_website_tech_stack	Technologies a website uses: CMS, ecommerce, analytics, CDN, email provider, frameworks	Website Tech Stack Detector
extract_pdf_text	Clean text or Markdown from public PDF and Word (.docx) URLs, with optional OCR	PDF Text Extractor
google_trends	Google Trends interest over time, regions and related queries, or what is trending now	Google Trends Scraper
get_run_results	Fetch or page through results of a run started by any tool above	–

Setup

You need an Apify account and your Apify API token. The server uses the Streamable HTTP transport at this endpoint:

https://artificially--business-data-mcp.apify.actor/mcp

Send your token as Authorization: Bearer <YOUR_APIFY_TOKEN>. The exact URL is also shown on the Actor's Endpoints tab in Apify Console.

Claude Code

claude mcp add --transport http business-data \
  https://artificially--business-data-mcp.apify.actor/mcp \
  --header "Authorization: Bearer <YOUR_APIFY_TOKEN>"

Cursor, VS Code, Windsurf and other MCP clients

{
  "mcpServers": {
    "business-data": {
      "url": "https://artificially--business-data-mcp.apify.actor/mcp",
      "headers": { "Authorization": "Bearer <YOUR_APIFY_TOKEN>" }
    }
  }
}

Claude (claude.ai and Claude Desktop) via Apify's MCP server

Claude's custom connectors sign in with OAuth, so connect through Apify's hosted MCP server, which loads this server's tools for you:

Settings → Connectors → Add custom connector, and paste:

https://mcp.apify.com?tools=artificially/business-data-mcp

The same URL works in any client that supports remote MCP servers with OAuth. Tools then appear with an artificially--business-data-mcp-- prefix.

Clients without remote HTTP support

Use the mcp-remote bridge:

{
  "mcpServers": {
    "business-data": {
      "command": "npx",
      "args": [
        "-y", "mcp-remote",
        "https://artificially--business-data-mcp.apify.actor/mcp",
        "--header", "Authorization: Bearer <YOUR_APIFY_TOKEN>"
      ]
    }
  }
}

Example prompts

• "Find open SAM.gov cybersecurity solicitations for NAICS 541512 set aside for small businesses, due in the next 30 days."
• "Who won the largest Department of Veterans Affairs cloud contracts since 2024, and for how much?"
• "List open German public tenders for software development with deadlines at least two weeks away."
• "Which engineering roles are Stripe, Notion and Airbnb hiring for in London right now?"
• "Search Google Jobs and LinkedIn for remote senior data engineer jobs posted this week."
• "Get the LinkedIn company pages of OpenAI and Anthropic and compare headcount and followers."
• "List YC W25 fintech startups that are hiring, with their founders."
• "Find the contact emails and LinkedIn pages for these 10 domains, then tell me which of them run Shopify."
• "Read this RFP PDF and summarize the evaluation criteria and deadlines."
• "Compare Google Trends interest for 'claude' and 'chatgpt' in the US over the last 12 months."

Pricing

You pay each underlying Actor's per-result price, and nothing extra for the MCP layer.

• Every tool call starts one run of the matching Artificially Actor on your Apify account, billed under that Actor's own pay-per-event pricing (for example, per company, per job or per document). Check each Actor's Store page (linked in the Tools table) for current prices.
• Tools use small default limits (usually 10-20 results) and pass them to the Actor, so a typical call costs cents.
• The MCP server itself runs in Apify Standby mode with 256 MB of memory. Its small platform usage while connected is billed to your account like any Actor run; it stops automatically when idle.
• Apify's free plan includes monthly credit that covers plenty of tool calls to try it out.

How it works

1. Your MCP client calls a tool, for example find_yc_companies with {"batches": ["W24"], "limit": 5}.
2. The server starts artificially/ycombinator-scraper with compactOutput: true and maxResults: 5, using your own Apify token, so the run appears in your Apify Console.
3. It waits for the run (up to 2-3 minutes depending on the tool, with progress notifications), then returns the dataset items as compact JSON:

{
  "actor": "artificially/ycombinator-scraper",
  "runId": "jGdWwQx1J5blfxiom",
  "status": "SUCCEEDED",
  "itemCount": 5,
  "statusMessage": "Saved 5 Y Combinator companies (248 match in the directory). Reached maxResults.",
  "items": [
    { "name": "Indemni", "website": "http://www.indemni.com", "oneLiner": "Cargo Theft and Fraud Prevention Platform", "batch": "Winter 2024", "status": "Active", "teamSize": 7, "isHiring": false }
  ]
}

If the run is not finished in time, the response contains the items saved so far, a note and the runId; the agent calls get_run_results to get the rest. Responses are capped at about 40,000 characters, with the remainder available through get_run_results and offset.

FAQ

Do I need to start the Actor manually? No. It runs in Standby mode: the first request starts it automatically and it shuts down after a period of inactivity. Starting it from the Console's Start button does nothing useful.

Can I use my own Actor inputs? Each tool exposes the most useful inputs. For full control, run the underlying Actor directly or add it to Apify's MCP server by name.

Is LinkedIn data scraped while logged in? No. Only public pages visible to logged-out visitors are read.

Where do I report issues or request tools? Open an issue on the Actor's Issues tab.