Data Legion MCP Server

What is Data Legion MCP Server?

The Data Legion MCP server lets an AI assistant enrich and search B2B person and company data from a conversation, using your Data Legion account. It is a hosted endpoint at https://api.datalegion.ai/mcp that Claude and other MCP clients connect to after you sign in.

Data Legion is a B2B data provider. Its person records match on an email, phone number, social profile URL, or a name plus context such as company or location, and its company records match on a domain, name, LinkedIn URL or ticker. Records carry a last verified date and a confidence score, so an assistant can tell you how fresh and how certain a match is.

How to use Data Legion MCP Server?

Prerequisites: a Data Legion account and an MCP client that supports remote Streamable HTTP servers with OAuth. According to Data Legion's Claude plugin, signing in requires the admin or developer role on your Data Legion account. Each successful match uses a credit from that account.

Server URL:

https://api.datalegion.ai/mcp

Claude Code

claude mcp add --transport http datalegion https://api.datalegion.ai/mcp

Run /mcp in Claude Code, select the server and sign in to Data Legion in your browser. Data Legion also publishes an official Claude plugin (datalegion-ai/datalegion-plugin on GitHub) that bundles this connection with four skills: enrich a list, find people, research a company, and report a data issue.

Claude Desktop

Go to Settings, then Connectors, choose Add custom connector and paste https://api.datalegion.ai/mcp. Click Connect and sign in with your Data Legion account. The same steps work on claude.ai.

Cursor

{
  "mcpServers": {
    "datalegion": {
      "url": "https://api.datalegion.ai/mcp"
    }
  }
}

Save this in ~/.cursor/mcp.json or a project's .cursor/mcp.json; Cursor opens the sign-in page in your browser.

VS Code

{
  "servers": {
    "datalegion": {
      "type": "http",
      "url": "https://api.datalegion.ai/mcp"
    }
  }
}

Save this in .vscode/mcp.json, start the server and complete the browser sign-in.

You can review or disconnect connected apps at https://www.datalegion.ai/dashboard/connected-apps.

Key features of Data Legion MCP Server

• Person enrichment from almost any identifier: email, phone, LinkedIn or other social URL, or a name with company, location, school or job title.
• Company enrichment from a domain, LinkedIn URL, ticker or name, with firmographics such as industry, headquarters, employee count and growth.
• SQL-style search over people and companies tables (for example SELECT * FROM people WHERE job_title ILIKE '%director%'), paged with limit and offset.
• Precision controls: include_fields to trim output, min_confidence for high-precision matches, and required_fields to skip rows that lack a needed field.
• Free input utilities (utility_clean, utility_validate) that normalize emails, phones and URLs before matching.
• Free data-issue reporting (report_person, report_company) for wrong, outdated, misattributed or duplicate records.
• Remote OAuth connection; nothing to install and no key in a config file.

Use cases of Data Legion MCP Server

• Filling in a contact list: "Enrich these 15 emails with job title, company and work email."
• Building a prospect list: "Find product managers at fintech companies in New York."
• Account research before outreach: "Give me a brief on stripe.com: size, growth and hiring."
• Cleaning a spreadsheet before import: "Validate and clean these phone numbers and LinkedIn URLs."
• Correcting bad data: "This person's job title is outdated; report it to Data Legion with the correct value."

Troubleshooting Data Legion MCP Server

• Sign-in is refused. Data Legion requires the admin or developer role to authorize the connector. Ask your workspace admin to grant the role.
• Too many credits used. Every successful match costs a credit. Ask the assistant to confirm row counts before large enrichments and to use include_fields and required_fields so only useful rows are returned.
• Search returns errors. Put paging in the limit and offset parameters, not as LIMIT/OFFSET inside the SQL.
• Low match rates. Run the free utility_clean or utility_validate tools on messy inputs first.