Guide

BetterKnocks MCP for AI assistants

Connect Cursor or Claude to BetterKnocks with MCP: log knocks, list leads, and read analytics from your workspace with natural language.

  • Field ops
  • D2D CRM
  • Knock logging
Create workspacePricing

MCP (Model Context Protocol) lets an AI client call tools against live data. BetterKnocks ships an MCP server so you can ask Cursor or Claude Desktop to list leads, log a knock, or pull conversion stats without copying CSV exports.

Requirements

  • Elite or Enterprise workspace
  • API key from workspace settings (same key as REST API)
  • npm install -g @betterknocks/mcp or npx at runtime

Setup in Cursor

  1. Create an API key in BetterKnocks workspace settings.
  2. Add an MCP server entry pointing at npx -y @betterknocks/mcp.
  3. Set BETTERKNOCKS_API_URL to https://betterknocks.com/api and BETTERKNOCKS_API_KEY to your key.
  4. Restart Cursor and confirm tools like list_leads and log_knock appear.

Tools you get

  • list_workspaces, list_leads, get_lead, create_lead, update_lead, update_lead_status
  • log_knock, log_activity
  • list_products, create_product
  • list_events, create_event
  • get_analytics

Example prompts that work

  • "List leads with status Interested from this week."
  • "Log a knock for 42 Oak St: nobody home."
  • "What is our knock-to-sit-down conversion this month?"
  • "Create a callback event tomorrow at 2pm for lead X."

MCP vs REST API

Use REST for production integrations (CRM sync, nightly ETL). Use MCP when a human plus AI is exploring data or doing ad hoc field updates from the desktop. Same API key, same workspace permissions.

Create workspacePricing