Skip to main content
Building your own AI agent with code or an agent framework? See Building Custom AI Agents (MCP) for integrating with custom MCP clients.
Connect AI assistants like Claude Desktop directly to your Credible semantic models using the Model Context Protocol (MCP). This integration enables natural language queries against your governed data layer—whether you’re analyzing data or using AI coding assistants to build applications.

​
Use Cases

Data Analysis: Ask your AI assistant questions about your data in natural language. The AI will automatically use Credible’s tools to query your semantic models and provide accurate answers. Coding Assistance: When building applications with Credible data (like React dashboards), your AI coding assistant can help generate queries and code using your semantic models.

​
Setup

​
Claude Desktop/Web

Requirements: Custom connectors require a paid Claude Desktop/Web subscription (Pro, Team, or Enterprise).
  1. Go to Settings > Connectors in your Claude Desktop/Web interface
  2. Click Add Custom Connector
  3. Enter a name (e.g., “Credible”)
  4. Enter your organization’s MCP URL: https://<your-org>.data.credibledata.com/mcp
  5. Click Add
  6. Click Connect and complete the authentication flow
  7. Start asking questions about your data!
Our MCP server exposes several tools for data analysis. We recommend starting with these two (and turning off the others):
  • suggestAnalysis - Get AI-powered suggestions for analyzing your data
  • executeQuery - Run queries against your semantic models
For complete technical details on all available tools and parameters, see the MCP Reference.

​
Next Steps

Building Custom AI Agents

Build your own AI applications and agents with your semantic models

MCP API Reference

Complete technical details of all MCP tools and parameters
⌘I