π This repository contains examples for the Microsoft Fabric MCP Server
Real-world examples demonstrating how to use the Microsoft Fabric Model Context Protocol (MCP) server to automate Fabric operations including lakehouse management, notebook execution, and data population workflows.
git clone https://github.com/santhoshravindran7/Fabric-Analytics-MCP.git
cd Fabric-Analytics-MCP
npm install && npm run build# Set your workspace ID (required)
export FABRIC_DEFAULT_WORKSPACE_ID="your-workspace-id-from-fabric-portal"
# Authenticate with Azure
az logingit clone https://github.com/your-username/Fabric-MCP-Testing.git
cd Fabric-MCP-Testing/examples/getting-started
node basic-demo.cjsExpected result: β Connection verified with 62 available MCP tools
| Directory | Purpose | Key Files |
|---|---|---|
getting-started/ |
New user introduction | basic-demo.cjs - Test your setup |
data-population/ |
Production workflows | execute-notebook-livy-corrected.cjs - 5K+ records |
troubleshooting/ |
Debug & investigation | test-capacity-issue.cjs - Reproduce bugs |
| File | Description |
|---|---|
SETUP.md |
Complete setup instructions |
docs/DEMO_SUCCESS_SUMMARY.md |
Detailed success stories and workflows |
| Directory | Purpose |
|---|---|
notebooks/ |
Jupyter notebooks for testing |
archive/ |
Historical scripts and experiments (local-only) |
- β MCP Server Connection: Establish and test connection
- β Workspace Access: Verify permissions and list items
- β Tool Discovery: Find all 62 available MCP functions
- π Data Population: Generate 5,000+ structured records
- π Lakehouse Creation: Full lakehouse with tables and data
- π Notebook Execution: Automated Spark job processing
- π Session Management: Livy API integration and monitoring
- π Issue Reproduction: Test cases for known issues
- π Investigation Scripts: Deep-dive into MCP behavior
β οΈ Error Handling: Robust error capture and reporting
- MCP Server: Fabric-Analytics-MCP
- Authentication: Azure CLI (
az login) - Platform: Microsoft Fabric / Power BI Premium
- APIs: Microsoft Fabric REST API + Livy Spark API
- Language: Node.js (CommonJS modules)
π Results Summary:
βββ π Lakehouse: Created with 3 tables
βββ π Records: 5,000+ structured entries
βββ β‘ Performance: Sub-minute execution
βββ π Repeatability: 100% success rate
π MCP Tools Available: 62 functions
βββ π Workspace Management: 5 tools
βββ π Lakehouse Operations: 8 tools
βββ π Notebook Control: 12 tools
βββ β‘ Spark Job Management: 15 tools
βββ π§ Utility Functions: 22 tools
- Learn MCP Integration: Real working examples
- Test Your Setup: Validation scripts included
- Production Patterns: Battle-tested workflows
- Automate Fabric: Lakehouse creation and data loading
- Spark Workflows: Notebook execution and monitoring
- ETL Pipelines: End-to-end data processing examples
- CI/CD Integration: Automated Fabric deployments
- Monitoring: Error handling and logging patterns
- Troubleshooting: Debug scripts for common issues
- Fabric-Analytics-MCP - The MCP server (required)
- Model Context Protocol - MCP specification
- Microsoft Fabric - Platform documentation
- Issue #12 - Missing capacity tools
- PR #9 - UUID session ID fix
- Issue #10 - Notebook content limitation
- Test scenarios welcome in
examples/troubleshooting/ - Production workflows needed in
examples/data-population/ - Documentation improvements always appreciated
- Bug reproductions help the main project
MIT License - Feel free to use these examples in your own projects.
- Start with:
examples/getting-started/basic-demo.cjs - Check:
SETUP.mdfor detailed instructions - Debug with: Scripts in
examples/troubleshooting/ - Ask questions: In the main Fabric-Analytics-MCP repository
β Star this repo if these examples helped you get started with Microsoft Fabric MCP automation!