Secoda MCP Server

Learn how to use the Model Context Protocol (MCP) to enable AI agents to securely access and interact with your Secoda data catalog, run SQL queries, and explore data lineage.

What is Model Context Protocol?

MCP is a protocol that enables AI tools and applications to connect with Secoda's data catalog and services in a secure, standardized way. It provides a structured method for AI models to:

• Search and retrieve data assets (tables, columns, dashboards, documentation)

• Execute SQL queries against your connected data warehouses

• Access data lineage and relationships between entities

• Retrieve business glossary terms and definitions

• Access Secoda documentation and knowledge articles

How MCP Works

Secoda hosts the MCP server as part of your workspace that provides access to your data catalog through a secure, authenticated interface.

When an AI tool or application needs to access Secoda data:

1. The tool connects to Secoda's MCP server endpoints

2. Authentication verifies the user's workspace and AI permissions

3. The tool can then access relevant Secoda data and functionality

4. All operations respect your existing data governance and access controls

Benefits of Using MCP

• Secure Access: All data access is authenticated and respects workspace permissions

• Live Data Queries: Execute SQL queries directly against your data warehouses

• Comprehensive Catalog Access: Search across all your data assets and documentation

• Data Lineage: Understand upstream and downstream dependencies

• Standardized Interface: Consistent interaction pattern across different AI tools

Available Tools

The Secoda MCP server provides the following tools:

Tool	Description
search_data_assets	Search for tables, columns, charts, and dashboards
search	General search across all Secoda resources including documentation
run_sql	Execute SQL queries against your connected data warehouses
retrieve_entity	Get detailed information about a specific data entity
entity_lineage	Explore upstream and downstream data dependencies
glossary	Access business terms and definitions from your glossary
get_secoda_docs	Retrieve Secoda documentation and knowledge articles
chart	Work with charts and visualizations

Setting things up

Secoda MCP

Secoda MCP is a custom server built with FastMCP that provides access to Secoda's data tools and functionalities.

Features

• Access to Secoda's data assets search

• Access to documentation search

• Ability to run SQL queries directly

• Integration with Secoda's data catalog

• Entity lineage visualization

• Glossary term retrieval

Getting Started

Prerequisites

• Python 3.8 or higher

• Access to Secoda's API

• API token for authentication with the Secoda API (can be generated at https://app.secoda.co/settings/api if you're an admin)

Installation

1. Clone the repository:

   git clone https://github.com/secoda/secoda-mcp.git
   cd secoda-mcp

2. Install dependencies:

   # Install runtime dependencies
   python -m pip install -r requirements.txt

Environment Variables

Variable	Description	Default
API_TOKEN	Required token for API authentication	None (Required)
API_URL	URL of the Secoda API endpoint	https://app.secoda.co/api/v1/

Instance URL Note

If you're using a regional instance or self-hosted deployment of Secoda, replace app.secoda.co with your specific instance URL:

• EU region: eu.secoda.co

• APAC region: apac.secoda.co

• Self-hosted: Your custom domain (e.g., secoda.yourcompany.com)

Integration with Tools

Using with Cursor

To integrate Secoda MCP with Cursor, add the following configuration to your ~/.cursor/mcp.json file:

{
    "mcpServers": {
        "secoda-mcp": {
            "command": "python",
            "args": [
                "/path/to/secoda-mcp/server.py"
            ],
            "env": {
                "API_TOKEN": "your-api-token",
                "API_URL": "https://app.secoda.co/api/v1/"
            }
        }
    }
}

Replace /path/to/secoda-mcp/server.py with the actual path to your server.py file, set your API token, and update the API_URL with your Secoda instance URL if not using app.secoda.co.

Using with Claude

To integrate Secoda MCP with Claude Desktop, add the following to your Claude Desktop configuration file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

Windows: %APPDATA%\Claude\claude_desktop_config.json

{
    "mcpServers": {
        "secoda-mcp": {
            "command": "python",
            "args": [
                "/path/to/secoda-mcp/server.py"
            ],
            "env": {
                "API_TOKEN": "your-api-token",
                "API_URL": "https://app.secoda.co/api/v1/"
            }
        }
    }
}

Replace /path/to/secoda-mcp/server.py with the actual path to your server.py file, set your API token, and update the API_URL with your specific Secoda instance URL if not using app.secoda.co. To access the Claude Desktop configuration file, open the Claude menu, go to "Settings", click on "Developer" in the left-hand bar, and then click on "Edit Config".

After updating the configuration, restart Claude Desktop to apply the changes.

Using with Other Tools

Secoda MCP can be integrated with various other tools:

• VS Code: Use the VS Code extension for MCP servers

• JetBrains IDEs: Install the MCP plugin from the marketplace

• Command Line: Access MCP functionalities directly via the CLI

• Web Applications: Embed MCP capabilities in web apps through the REST API

Troubleshooting

• If you encounter connection issues, ensure Secoda's API is running at the configured URL

• If you get authorization errors, verify your API token is correct and properly set as an environment variable

• For permission errors, check your authentication settings in your Secoda account

License

This project is licensed under the MIT License - see the LICENSE file for details.

Security

If you discover any security related issues, please email [email protected] instead of using the issue tracker.

FAQ

Why am I getting authentication errors? Ensure you have the correct permissions in your Secoda workspace and that MCP is enabled by your administrator.

I'm seeing connection timeouts MCP connections can be sensitive to network conditions. Try restarting your AI client or check with your network administrator about proxy settings.