MCP Server

SignalSmith exposes its capabilities through a Model Context Protocol (MCP) server, enabling external AI tools to interact with your CDP programmatically.

Overview

The MCP server runs alongside the control plane and provides a standardized interface for AI agents to discover and use SignalSmith tools. This allows integration with Claude Desktop, Cursor, and any other MCP-compatible client.

Available Tools

Schema Tools

Tool	Description
`list_models`	List all models in the workspace
`get_model`	Get model details including SQL and columns
`list_relationships`	List entity relationships
`list_traits`	List trait definitions

Audience Tools

Tool	Description
`list_audiences`	List all audiences
`get_audience`	Get audience details including filter conditions
`estimate_audience_size`	Estimate audience member count
`validate_filter_node`	Validate a filter condition
`compile_filter`	Compile filter to SQL

Sync Tools

Tool	Description
`list_syncs`	List all syncs
`list_audience_syncs`	List audience syncs

Destination Tools

Tool	Description
`list_destinations`	List configured destinations
`list_sources`	List connected sources

Source Introspection Tools

Tool	Description
`introspect_source`	Discover schemas in a source
`list_tables`	List tables in a schema
`list_columns`	List columns in a table

Write Tools

Tool	Description
`create_audience`	Create a new audience
`update_audience`	Modify audience filters
`create_trait`	Create a new trait
`create_destination`	Configure a new destination
`trigger_sync`	Trigger a sync run
`create_relationship`	Create entity relationship

Journey Tools

Tool	Description
`list_journeys`	List all journeys
`get_journey`	Get journey details
`validate_journey`	Validate journey configuration
`trigger_journey`	Trigger a journey run

Authentication

The MCP server authenticates requests using API keys:

Create an API key in Settings > API Keys
Configure your MCP client with the key

Integration with Claude Desktop

Add to your Claude Desktop configuration:

{
  "mcpServers": {
    "signalsmith": {
      "url": "http://localhost:8090",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY",
        "X-Workspace-ID": "YOUR_WORKSPACE_ID"
      }
    }
  }
}

Available Resources

The MCP server also exposes resources that provide context to AI agents:

Schema specifications — Entity type and relationship schemas
Filter language spec — Documentation of the audience filter syntax
Trait definitions — Available traits and their types

AI Agent Guardrails