Connect any MCP-compatible agent to Datasigma and let it search contacts, reveal work emails and look up companies on demand — every record carrying the same source URL you see in-app.
An agent asked to find a decision-maker will happily invent one. Training data goes stale, there's no way to verify what it returns, and wiring up a custom tool for every data source is work you have to maintain forever.
Point your MCP client at Datasigma and the agent discovers the tools automatically. When it needs a person, it searches — and every result comes back with its source.
# The agent calls the search_contacts tool
{
"tool": "search_contacts",
"arguments": { "title": "Head of Sales", "seniority": "director" }
}
# → tool result
{
"results": [
{
"name": "Jordan Lee",
"title": "Head of Sales",
"company": "Acme Inc.",
"source_url": "https://acme.com/team"
}
]
}
Point your MCP client — Claude, Cursor or your own agent — at Datasigma with your API key.
It automatically discovers search_contacts, reveal_contact and lookup_company.
When the agent needs a person or a company, it queries live data instead of guessing.
Each record returns the public page it came from, so the agent's output stays verifiable.
search_contactsFind people by name, title, seniority or domain, straight from the live contact index.
reveal_contactPull a verified work email or phone number on demand — only when the agent actually needs it.
lookup_companyResolve a company by name or domain to a structured record the agent can reason over.
Each result carries the public page it came from — provenance your agent can quote in its answer.
Claude, Cursor or your own framework — anything that speaks the Model Context Protocol.
Agents get the exact records you see in the dashboard — not a separate, lower-quality feed.
The Datasigma MCP server is in private beta. Join the waitlist and we'll reach out with access.
Join the beta waitlist