Bonnard Analytics

v1.0.0Analyticsstable

Open-source agentic schema for reliable data outputs. Query data through MCP and via our SDK. Create apps, embed data or just simply explore through your preferred agent.

agenticagentic-analyticsagentic-schemaai-analyticsbigquery
Share:
26
Stars
0
Downloads
0
Weekly
0/5

What is Bonnard Analytics?

Bonnard Analytics is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to open-source agentic schema for reliable data outputs. query data through mcp and via our sdk. create apps, embed data or just simply explore through your preferred agent.

Open-source agentic schema for reliable data outputs. Query data through MCP and via our SDK. Create apps, embed data or just simply explore through your preferred agent.

This server falls under the Analytics category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Open-source agentic schema for reliable data outputs. Query

Use Cases

Agentic schema for data
dbt and BigQuery integration
Data modeling
bonnard-data

Maintainer

LicenseApache 2.0
Languagetypescript
Versionv1.0.0
UpdatedApr 27, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx bonnard

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Bonnard Analytics

Bonnard is an open-source, self-hosted agentic semantic layer that sits between AI agents and your databases, providing reliable, governed data access through a cubes-and-views schema model. It supports PostgreSQL, DuckDB, Snowflake, BigQuery, Databricks, Redshift, and ClickHouse, and exposes an MCP server so Claude and other AI agents can query structured business metrics without writing raw SQL. Data teams building on dbt or maintaining a metrics layer can deploy Bonnard via Docker and give AI agents safe, schema-governed access to production data while maintaining full control over what can be queried.

Prerequisites

  • Node.js 20+ and Docker with Docker Compose installed
  • A supported database: PostgreSQL, DuckDB, Snowflake, BigQuery, Databricks, Redshift, or ClickHouse
  • Database connection credentials (host, port, database name, username, password)
  • npx available to run the @bonnard/cli scaffolding tool
  • An MCP-compatible client such as Claude Desktop or Cursor
1

Scaffold a new Bonnard project

Use the Bonnard CLI to generate a new self-hosted project directory with Docker Compose configuration and example cube files.

npx @bonnard/cli init --self-hosted
cd bonnard-project
2

Configure your database connection

Edit the generated .env file to specify your database type and connection credentials using the CUBEJS_DB_* environment variables.

# .env example for PostgreSQL:
CUBEJS_DB_TYPE=postgres
CUBEJS_DB_HOST=localhost
CUBEJS_DB_PORT=5432
CUBEJS_DB_NAME=analytics
CUBEJS_DB_USER=your_db_user
CUBEJS_DB_PASS=your_db_password
CUBEJS_API_SECRET=auto_generated_on_init
ADMIN_TOKEN=your_chosen_admin_bearer_token
3

Start the Bonnard server

Launch the Bonnard server and supporting services using Docker Compose.

docker compose up -d
4

Define your schema with cube and view YAML files

Add cube definitions to bonnard/cubes/ and view definitions to bonnard/views/ to declare the metrics and dimensions your agents can query.

5

Deploy your schema

Push schema changes to the running server without a container restart using the Bonnard CLI.

bon deploy
6

Connect your MCP client

Add Bonnard to your Claude Desktop or Cursor MCP configuration, pointing to your Bonnard server URL with an authorization token.

{
  "mcpServers": {
    "bonnard": {
      "url": "http://localhost:3000/mcp",
      "headers": {
        "Authorization": "Bearer your_admin_token_here"
      }
    }
  }
}

Bonnard Analytics Examples

Client configuration

Connect Claude Desktop to a locally running Bonnard server using HTTP MCP transport with bearer token authentication.

{
  "mcpServers": {
    "bonnard": {
      "url": "http://localhost:3000/mcp",
      "headers": {
        "Authorization": "Bearer your_admin_token_here"
      }
    }
  }
}

Prompts to try

Sample queries that leverage the semantic layer to retrieve business metrics without writing raw SQL.

- "What is our total revenue by product category for Q1 2025?"
- "Show me the top 10 customers by lifetime value"
- "What are the daily active user counts for the past 30 days?"
- "Compare conversion rates across marketing channels for the last quarter"
- "List all available metrics and dimensions in the Bonnard schema"
- "What is the average order value broken down by region and month?"

Troubleshooting Bonnard Analytics

Docker Compose fails to start with database connection errors

Verify your CUBEJS_DB_* environment variables in the .env file match your database's actual host, port, and credentials. If connecting to a local database from Docker, use 'host.docker.internal' instead of 'localhost' for the DB_HOST value.

MCP client receives 401 Unauthorized responses

Ensure the ADMIN_TOKEN in your .env file matches the Bearer token in your MCP client's Authorization header exactly. If ADMIN_TOKEN is not set, the MCP endpoint is unprotected — add a token for any non-local deployment.

Schema changes are not reflected after editing cube YAML files

Run `bon deploy` from the project root to push updated cube and view definitions to the running server. Schema changes are not automatically hot-reloaded; `bon deploy` is required to apply them without restarting containers.

Frequently Asked Questions about Bonnard Analytics

What is Bonnard Analytics?

Bonnard Analytics is a Model Context Protocol (MCP) server that open-source agentic schema for reliable data outputs. query data through mcp and via our sdk. create apps, embed data or just simply explore through your preferred agent. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Bonnard Analytics?

Follow the installation instructions on the Bonnard Analytics GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Bonnard Analytics?

Bonnard Analytics works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Bonnard Analytics free to use?

Yes, Bonnard Analytics is open source and available under the Apache 2.0 license. You can use it freely in both personal and commercial projects.

Browse More Analytics MCP Servers

Explore all analytics servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "bonnard": { "command": "npx", "args": ["-y", "bonnard"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Bonnard Analytics?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides