Getting Started with the Solid MCP Server

What You Can Do

Once connected, you can:

• Build agentic workflows that query your data automatically as part of a larger AI pipeline
• Generate SQL from plain English — describe what you want and get the query back instantly
• Integrate Solid's data layer directly into the AI tools and agents you build
• Explore your data interactively — tables, columns, metrics, and more through conversation

Before You Start

You will need:

1. A data warehouse connection configured in Solid — the text2sql tool generates SQL that runs against your DWH. Ask your Solid admin to confirm your connection is set up before you begin.

2. The Solid MCP Server URL:

   https://mcp.production.soliddata.io/mcp

3. A way to log in — either via your browser (easiest) or an access token from your Solid admin (for automated tools).

Which Setup Is Right for You?

Option A — Sign in with your browser

Easiest option. No tokens needed.

Setting up in Cursor

1. Open Settings → MCP (or press Cmd+Shift+P on Mac / Ctrl+Shift+P on Windows and search for "MCP: Add Server")

2. Click + Add new MCP server and enter:

   • Name: Solid
   • Type: streamable-http
   • URL: https://mcp.production.soliddata.io/mcp

3. Save. The next time you use Solid, Cursor will open a browser window — just log in with your Solid credentials.

4. That is it! You can now start generating SQL queries directly in the chat.

Prefer to configure it manually? Add this to your .cursor/mcp.json:

{
  "mcpServers": {
    "Solid": {
      "url": "https://mcp.production.soliddata.io/mcp",
      "transport": "streamable-http"
    }
  }
}

Setting up in other tools (Claude Desktop, Windsurf, etc.)

Add a new MCP server with:

• URL: https://mcp.production.soliddata.io/mcp
• Transport: streamable-http

Sign in when prompted. The tool handles token refresh automatically — no manual steps needed.

Option B — Use an access token

For CrewAI, Workato, scripts, and any automated environment.

Step 1 — Get your token

Ask your Solid admin for a management key, then run this command to exchange it for an access token (JWT):

curl --location 'https://backend.production.soliddata.io/api/v1/auth/exchange_user_access_key' \
  --header 'Content-Type: application/json' \
  --data '{
    "management_key": "YOUR_MANAGEMENT_KEY"
  }'

You will get back a token that looks like this:

"eyJhbGciOiJSUzI1NiIsInR5cCI6IkpXVCIs..."

Note: Tokens expire after 30 minutes. Make sure your application refreshes the token every 25–30 minutes by calling the exchange endpoint again with your management key.

Step 2 — Connect your tool

Pass the token as a Bearer header when connecting to the Solid MCP Server.

CrewAI example:

from crewai.tools import MCPServerHTTP

mcp_server = MCPServerHTTP(
    url="https://mcp.production.soliddata.io/mcp",
    headers={"Authorization": "Bearer YOUR_TOKEN_HERE"},
    streamable=True,
    cache_tools_list=True,
)

Any other HTTP-based tool:

Tip: Save your token as an environment variable (e.g. SOLID_MCP_TOKEN) instead of pasting it directly into your code.

Option C — Connect via Microsoft Copilot Studio

For Microsoft Copilot Studio agents running in Teams or other Copilot channels.

This option uses the native MCP onboarding wizard in Copilot Studio to connect directly to the Solid MCP Server — no bridge or additional infrastructure needed.

Before you start

You will need:

• A Solid management key from your Solid admin
• Your Semantic layer ID(s) — the UUID(s) of your Solid semantic layer(s)
• Access to Microsoft Copilot Studio

Setup steps

1. In Copilot Studio, open your agent and go to Tools.
2. Select Add a tool → New tool → Model Context Protocol.
3. Enter:
   • Server name: Solid Text2SQL
   • Server description: Converts natural language questions to SQL using Solid's semantic layer
   • Server URL: https://mcp.production.soliddata.io/mcp
4. Choose your authentication method:
   • API key — if your Solid admin has provided one
   • OAuth 2.0 — if your organization uses OAuth
5. Select Create, then Add to agent.

Your Copilot Studio agent can now call the Solid text2sql tool at runtime.

Configuring your agent

Once connected, configure your agent to route data questions to Solid:

• Agent instructions — Tell the agent when to use the Solid tool, for example: "For data or SQL questions, use the Solid Text2SQL action."
• Web search — Consider limiting web search for data questions so the agent prefers the Solid tool.
• Test — Use "Show activity map when testing" to confirm the tool is being invoked and the SQL is returned correctly.

Troubleshooting

Quick Reference

Available Tools

Examples

Text-to-SQL — single data source

"What were total sales last month?"

The tool generates the SQL query. You then run it against your data warehouse.

Text-to-SQL — multiple data sources

"Show me top 10 customers by revenue."

If you have multiple source systems, the tool returns a list of options. Reply with your choice (e.g. snowflake), then call again with source_system set to receive the SQL query.

Security

• Encryption — All traffic is encrypted via HTTPS (TLS 1.2 or later).
• Authentication — Browser sign-in uses OAuth 2.1 through your organization's identity provider. Tokens are short-lived and automatically refreshed by the client.
• Data access — Access to the actual data is controlled by your data warehouse. The SQL generated by the Solid MCP Server runs under whatever credentials you have configured in your DWH connection.

How It Works

1. Add the Solid MCP Server URL to your MCP-compatible tool and complete sign-in.
2. The Solid MCP Server's tools are discovered automatically over the MCP protocol.
3. When a question or query is triggered — by you or by an agent — it is sent to the server along with your auth token.
4. The server validates your token and routes your request:
   • text2sql — generates SQL and optionally prompts for a source system selection
5. The SQL is returned to your tool or agent to run against your data warehouse and act on, display, or pass along to the next step in your workflow.