Your agent alreadyknows how to use this.
188M+ person profiles and 71M+ company profiles, callable from Claude, Cursor, ChatGPT, Gemini, OpenClaw, and any MCP client. No glue code, no parsing: your agent calls it like a native tool.
Drop into your agent in two minutes
Claude, ChatGPT, Codex and Grok connect with your Data Legion sign-in, so there is no key to paste. Claude Code, Cursor, Gemini CLI and other coding tools take one block in their MCP config, like the .mcp.json here. Same API key, same credits.
{
"mcpServers": {
"datalegion": {
"type": "http",
"url": "https://api.datalegion.ai/mcp",
"headers": {
"API-Key": "YOUR_API_KEY"
}
}
}
}Native MCP tool surface
Nine tools across person, company, utility, and reporting: exactly what your agent calls.
person_enrich
Look up a person by email, phone, name, LinkedIn, or other identifiers.
person_search
Query people using SQL WHERE syntax over 100+ data points.
company_enrich
Look up a company by domain, name, ticker, or LinkedIn URL.
company_search
Query companies using SQL WHERE syntax over 50+ firmographics.
utility_clean
Normalize emails, phones, names, and domains. No credit cost.
utility_hash_email
Hash an email (SHA-256, SHA-1, MD5) for privacy-safe matching.
utility_validate
Validate data format and surface cleaning suggestions.
report_person
Flag a wrong, outdated, or misattributed value in a person record. No credit cost.
report_company
Flag a wrong, outdated, or misattributed value in a company record. No credit cost.
Enrich, search and utility tools are all available self-serve.
$ brew install datalegion-ai/tap/datalegion-cli
$ datalegion-cli login
# opens a sign-in link; confirm the code it shows
$ datalegion-cli person enrich \
--email jane.doe@example.com
{
"full_name": "Jane Doe",
"job_title": "VP of Engineering",
"company_name": "Acme Corp",
"city": "San Francisco",
...
}A native shell tool for AI coding agents
Install with Homebrew or pip. Claude Code, Cursor, OpenClaw, Cline, and Aider can call it as a shell tool with no extra wiring. JSON output to stdout, logs to stderr, built for scripts, pipelines, and agent loops.
CLI Setup GuideWhat agents build with our data
Use cases that string our tools together. Paste a prompt into any assistant connected to Data Legion.
AI SDR enrichment loop
Hand your AI SDR a contact list. It enriches each row, scores fit against your ICP, drafts a personalized opener, and asks before sending.
Use Data Legion person enrich to look up each of these contacts by email: [paste emails]. For each match, list their name, title, company and LinkedIn URL. Flag the ones that fit this profile: [describe your ICP]. Draft a two-sentence opener for each fit and show me the drafts before sending anything.Recruiter shortlist agent
Recruiter agent reads a JD, calls person_search to build a shortlist, calls person_enrich for contact info, and drafts personalized outreach.
Read this job description: [paste JD]. Use Data Legion person search to find 25 people who match the role in [city, state]. Enrich the 10 strongest matches for contact details and draft a short, personal outreach note for each that references their current role.Org-mapping research agent
Research agent reads a company name, calls company_enrich for firmographics, then person_search to map the org chart by seniority and function.
Use Data Legion company enrich to look up [company domain] and summarize its size, industry and locations. Then use person search to list its directors and above, grouped by function and seniority, with each person's title.Person tools for your agents
Every product is a tool your agent can call
Person Enrichment API
Person lookups for your agents. Match by email, phone, LinkedIn, or name and get structured profiles back.
Person Search API
Build targeted lists with SQL queries. Power agent workflows with on-demand person search.
Person Discovery APIBeta
Find people using natural language queries. Describe who you’re looking for and get matched results.
Person Data License
Bulk data for RAG pipelines, fine-tuning, and vector stores. CSV, JSON, Parquet, or Delta with monthly refreshes.
Company tools for your agents
Company intelligence for account scoring, market mapping, and research agents
Company Enrichment API
Match and append company data to your records. 6 matching methods with confidence levels.
Company Search API
Query companies on demand with SQL-like filters. Up to 100 results per query.
Company Discovery APIBeta
Find companies using natural language queries. Describe what you're looking for and get matched results.
Company Data License
Bulk company data feeds delivered to your cloud storage. CSV, JSON, Parquet, or Delta with monthly refreshes.
Common Questions
What agent builders ask us most
Yes. The MCP server is a thin layer over the same REST API. The same API key, credits, and rate limits apply. Tool calls return structured JSON that the agent can parse and act on.
Tool calls use credits exactly as the matching REST endpoints do: one credit for each record returned, so a lookup with no match costs nothing. Self-serve credits cover the enrichment and search tools, and the utility and reporting tools cost no credits.
In Claude, ChatGPT, Claude Code, Codex and Grok, add https://api.datalegion.ai/mcp as a connector and sign in with your Data Legion account; no key to paste. In other clients, add the server URL and your API-Key header to the MCP configuration. We have setup steps for Claude Code, Claude Desktop, Cursor, Gemini, ChatGPT, Codex, Grok, and OpenClaw in the docs.
Yes. An agent can sign in on your behalf by following https://www.datalegion.ai/auth.md: you approve it by reading back a short code, and it gets its own API key on the team you choose. Every agent you approve is listed on your Connected Apps page, where you can disconnect it.
The CLI is fully async, outputs JSON to stdout (logs to stderr), and is built to be called as a shell tool by AI coding agents. Claude Code, Cursor, OpenClaw, Cline, and Aider can run it directly. Install with brew or pip.
We expose discovery via /.well-known/mcp/server-card.json (SEP-1649) and /.well-known/mcp.json. Agents that support these patterns can detect our server, its tools, and its auth method automatically.
Anything that speaks Model Context Protocol. We ship configs for Claude Code, Claude Desktop, ChatGPT, Codex, Grok, Gemini, Cursor, OpenClaw, Cline, Continue, Windsurf, and Goose. New MCP-compatible clients work out of the box.
Connect your agent in 60 seconds
Same API key works across REST, MCP, and CLI. Self-serve credits or enterprise agreements available.