Skip to main content
Landbase MCP connects an AI assistant to Landbase’s company and people data. It exposes tools for building prospect lists, inspecting datasets, matching records, and running enrichment, qualification, and research workflows. The remote endpoint is https://mcp.landbase.com/mcp. Connect with a client that supports Streamable HTTP and OAuth. See Connect an MCP client.

​
What you can do

  • Build a company or people dataset from a natural-language search.
  • Inspect field names, preview rows, and trace dataset lineage.
  • Upload a small CSV or JSONL dataset.
  • Match companies and people to Landbase records.
  • Enrich, qualify, research, or transform a dataset.
  • Publish a dataset and request a download link.
See Supported tools for parameters and examples.

​
Your first workflow

Ask your assistant:
Find B2B SaaS companies in Texas with 50–200 employees. Show me the available accounts first, then preview the results before running any paid enrichment.
If you have several accounts, choose one and have the assistant use its account_id explicitly on every account-scoped call, including polling and confirmation calls. The server does not retain the selection between calls. Searches and workflows can take longer than one tool call; continue polling the returned handle or run ID rather than starting duplicate work. Review credit quotes or row counts before approving paid work. Workflow and download tools use confirm: true; research search uses a confirmation turn in the original session. See Limits and credits. Try Build your first prospect list for a complete search-to-download walkthrough. If you use Grok Bot, start with Connect Landbase to Grok Bot.