rcandelu/fabric-mcp-server

Fabric MCP Server

🚀 Remote MCP Server for Microsoft Fabric with BI capabilities - Integrates with OneLake semantic models and provides read-only query access with insights management.

🎯 Features

• Microsoft Fabric Integration: Direct connection to OneLake semantic models
• Read-Only SQL Access: Secure query execution with validation
• BI Insights Management: Capture and store analysis insights
• Pre-built Prompts: Sales analysis and BI report generation
• Azure Functions Hosting: Serverless architecture for scalability
• Claude Desktop Compatible: Works seamlessly with Claude AI

🏗️ Architecture

Claude Desktop <-> Azure Functions (MCP Server) <-> Microsoft Fabric
                            |
                            v
                   Azure Blob Storage
                    (Insights Memo)

🛠️ Tools Available

1. list_tables

List all available tables in your Fabric lakehouse.

2. read_query

Execute read-only SQL queries on your Fabric data with:

• Query validation (SELECT only)
• 30-second timeout protection
• Formatted JSON response

3. append_insight

Save important findings to the company insights memo:

• Categorized insights (general, financial, operational, marketing)
• Tagging system for easy retrieval
• Persistent storage in Azure Blob

📝 Resources

insights-memo

Markdown document containing all captured insights:

• Organized by category
• Timestamped entries
• Full history of analysis findings

💡 Prompts

analyze-sales-data

Comprehensive sales analysis including:

• Monthly trends
• Top performers
• YoY comparisons
• Category breakdowns

generate-bi-report

Customizable BI reports:

• Executive dashboards
• Operational efficiency
• Financial analysis
• Marketing performance

🚀 Deployment

Prerequisites

1. Azure Subscription
2. Microsoft Fabric workspace with lakehouse
3. Azure Storage Account
4. Azure Functions Core Tools (for local development)

Setup Steps

1. Clone the repository

   git clone https://github.com/rcandelu/fabric-mcp-server.git
   cd fabric-mcp-server
   

2. Configure Azure Resources

   Create an Azure Function App:

   az functionapp create \
     --resource-group myResourceGroup \
     --consumption-plan-location westeurope \
     --runtime python \
     --runtime-version 3.11 \
     --functions-version 4 \
     --name fabric-mcp-server \
     --storage-account mystorageaccount
   

3. Set Environment Variables

   In Azure Portal, add these Application Settings:

   • FABRIC_TENANT_ID: Your Azure AD tenant ID
   • FABRIC_CLIENT_ID: Service principal client ID
   • FABRIC_CLIENT_SECRET: Service principal secret
   • FABRIC_WORKSPACE_ID: Fabric workspace ID
   • FABRIC_LAKEHOUSE_ID: Lakehouse ID
   • STORAGE_ACCOUNT_NAME: Azure Storage account name
   • STORAGE_ACCOUNT_KEY: Storage account key

4. Configure GitHub Secrets

   Add these secrets to your GitHub repository:

   • AZURE_CREDENTIALS: Service principal JSON
   • AZURE_FUNCTIONAPP_PUBLISH_PROFILE: Download from Azure Portal

5. Deploy

   Push to main branch to trigger automatic deployment:

   git push origin main
   

💻 Local Development

1. Install dependencies

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate
   pip install -r requirements.txt
   

2. Configure local settings

   Copy local.settings.json and fill in your values.

3. Run locally

   func start
   

🔧 Claude Desktop Configuration

Add to your Claude Desktop config:

{
  "mcpServers": {
    "fabric-bi": {
      "url": "https://fabric-mcp-server.azurewebsites.net/mcp",
      "headers": {
        "x-api-key": "your-api-key"
      },
      "capabilities": {
        "tools": true,
        "resources": true,
        "prompts": true
      }
    }
  }
}

🔐 Security Considerations

• Read-Only Access: Only SELECT queries are allowed
• Query Validation: Dangerous SQL keywords are blocked
• Timeout Protection: 30-second query timeout
• Authentication: Service principal with minimal permissions
• API Key: Additional layer of security for endpoints

📊 Usage Examples

List Available Tables

User: Show me what tables are available in our Fabric lakehouse
Claude: I'll list all the available tables for you...
[Uses list_tables tool]

Execute Analysis Query

User: What were our top 10 products by revenue last quarter?
Claude: I'll query the sales data to find your top performing products...
[Uses read_query tool with appropriate SQL]

Save Insights

User: That's interesting, please save these findings
Claude: I'll save these insights to your company memo...
[Uses append_insight tool]

🤝 Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

📝 License

MIT License - see LICENSE file for details

🔗 Resources

• Microsoft Fabric Documentation
• MCP Protocol Specification
• FastMCP Framework
• Azure Functions Python Guide

👥 Support

For issues and questions:

• Open an issue on GitHub
• Check existing issues for solutions
• Consult Microsoft Fabric and Azure documentation

Built with ❤️ for the Microsoft Fabric and MCP communities