MCP Integration

Connect AI coding tools directly to LocalStitch via the Model Context Protocol (MCP). Search businesses, manage lists, and access contacts through natural language.

What is MCP?

The Model Context Protocol is an open standard that lets AI assistants interact with external tools and data sources. Instead of copying and pasting data, you can ask your AI assistant to work with LocalStitch directly.

Without MCP

Copy data from LocalStitch → Paste into AI → Copy results → Paste back

With MCP

"Find coffee shops in Austin" → Done

Supported Clients

Claude Code
Cursor
Windsurf
VS Code

Any MCP-compatible client will work with LocalStitch's MCP server.

Setup Instructions

Claude Code

Run this command in your terminal:

bash
claude mcp add localstitch \
  -"text-cyan-400">-transport streamable"text-cyan-400">-http \
  -"text-cyan-400">-url "https://localstitch.co/mcp"

You'll be prompted to authenticate with your LocalStitch account via OAuth.

Cursor

Add this to your .cursor/mcp.json:

json
{
  "mcpServers": {
    "localstitch": {
      "url": "https://localstitch.co/mcp",
      "transport": "streamable-http"
    }
  }
}

Other MCP clients

For Windsurf, VS Code, and other clients, visit the API Console MCP tab for client-specific setup instructions.

Authentication

MCP uses OAuth 2.0 for authentication — no API key required. When you first connect, you'll be redirected to LocalStitch to authorize the connection.

Secure — Uses OAuth 2.0 PKCE flow

Scoped — Access limited to your workspace

Revocable — Disconnect anytime from workspace settings

Available Tools

ToolDescription
search_businessesSearch for businesses with filters (name, category, location, etc.)
list_listsGet all lists in your workspace
get_listGet details for a specific list
get_list_itemsGet businesses in a list (paginated)
create_listCreate a new list
add_to_listAdd a business to a list
list_contactsList enriched contacts with filtering
get_contactGet details for a specific contact

Enrichment not available via MCP

Credit-consuming operations (business enrichment) are not available through MCP. Use the REST API or LocalStitch UI for enrichment.

Example Prompts

Once connected, try these natural language queries:

"Find Italian restaurants in San Francisco with at least 4 stars"
"Create a list called 'Q1 Prospects'"
"Add this business to my leads list"
"Show me coffee shops in Austin with email addresses"
"How many contacts do I have from enriched businesses?"
"Get businesses in my 'High Priority' list"

Tool Schema Example

Here's what a search tool call looks like under the hood:

json
// Example MCP tool call
{
  "name": "search_businesses",
  "arguments": {
    "query": "coffee shop",
    "filters": {
      "state": "California",
      "city": "San Francisco",
      "has_email": true,
      "rating_min": 4.0
    },
    "limit": 10
  }
}

Rate Limits

MCP requests share the same rate limit as the REST API:

Combined limit60 requests/minute

Plan accordingly if using both MCP and REST API simultaneously.

llms.txt

For AI tools that support it, LocalStitch provides a llms.txt endpoint with full API documentation in a format optimized for LLM consumption:

https://localstitch.co/api/llms.txt