Excel-Power-Pivot-MCP
A Model Context Protocol (MCP) server that enables AI assistants to interact with Excel Power Pivot data models. Create and manage DAX measures, relationships, and more through natural language.
README
Excel Power Pivot MCP Server
A Model Context Protocol (MCP) server that enables AI assistants to interact with Excel Power Pivot data models. Create and manage DAX measures, relationships, and more through natural language.
Features
- Workbook Discovery & Connection - Discover and connect to open Excel workbooks with Power Pivot data models
- DAX Query Execution - Run DAX queries and preview table data
- Measure Management - Create, update, and delete DAX measures with auto-formatting
- Relationship Management - Create, delete, and activate/deactivate table relationships
- Model Exploration - List tables, columns, measures, relationships, hierarchies, KPIs, and dependencies
- Data Profiling - Analyze column statistics (min, max, distinct count, nulls, sample values)
- Power Query Discovery - List Power Queries (M code) in the workbook
- Table Management - Add Excel tables to the data model, refresh tables/model
[!WARNING] Use at your own risk. This tool modifies your Excel Power Pivot data models directly. The author is not responsible for any data loss, corruption, or damage to your workbooks. Always maintain backups of your Excel files before using this tool.
Requirements
- Windows 10/11 (required for Excel COM interop)
- Microsoft Excel 2013+ with Power Pivot enabled
Installation
Download the latest ExcelPowerPivotMcp.exe from the Releases page.
No installation required - just download and configure your MCP client.
MCP Client Configuration
Claude Desktop
Add to your claude_desktop_config.json:
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"excel-powerpivot": {
"command": "C:/path/to/ExcelPowerPivotMcp.exe"
}
}
}
Cursor
Add to your Cursor MCP settings:
{
"mcpServers": {
"excel-powerpivot": {
"command": "C:/path/to/ExcelPowerPivotMcp.exe"
}
}
}
Antigravity (VS Code)
Add to your .vscode/mcp.json:
{
"servers": {
"excel-powerpivot": {
"type": "stdio",
"command": "C:/path/to/ExcelPowerPivotMcp.exe"
}
}
}
Usage
1. Open Excel with Power Pivot
Open your Excel workbook that contains a Power Pivot data model.
2. Connect via MCP
The AI assistant will first discover and connect to your workbook:
AI: Let me discover your Excel workbooks...
→ discover_workbooks()
→ Found: MyModel.xlsx (hasDataModel: true)
AI: Connecting to your workbook...
→ connect_workbook(workbook_name: "MyModel.xlsx")
→ Connected!
3. Explore and Modify
AI: Let me see what's in your data model...
→ get_model_summary()
→ 5 tables, 12 measures, 6 relationships
AI: I'll create a new measure for you...
→ create_measure(
tableName: "Sales",
measureName: "Total Revenue",
expression: "SUM(Sales[Amount])"
)
AI: Don't forget to save!
→ save_workbook()
Available Tools
Connection
| Tool | Description |
|---|---|
discover_workbooks |
Find open Excel workbooks with Power Pivot models |
connect_workbook |
Connect to a specific workbook |
get_connection_status |
Check current connection status |
save_workbook |
Save the connected workbook |
refresh_model |
Refresh the entire data model |
Model Metadata
| Tool | Description |
|---|---|
get_model_summary |
Comprehensive model overview |
list_tables |
List all tables with row counts |
list_columns |
List columns in a table |
list_measures |
List measures with expressions |
list_relationships |
Show table relationships |
list_hierarchies |
List user-defined hierarchies |
list_kpis |
List Key Performance Indicators |
get_dependencies |
Show calculation dependencies |
list_power_queries |
List Power Queries (M code) |
list_excel_tables |
List Excel tables (ListObjects) |
DAX Queries
| Tool | Description |
|---|---|
run_dax |
Execute DAX queries or preview table data |
analyze_column |
Get column statistics and sample values |
Measure CRUD
| Tool | Description |
|---|---|
create_measure |
Create a new DAX measure |
update_measure |
Update expression/name/description |
delete_measure |
Delete a measure |
Relationship CRUD
| Tool | Description |
|---|---|
create_relationship |
Create a table relationship |
delete_relationship |
Delete a relationship |
set_relationship_active |
Activate/deactivate a relationship |
Table Operations
| Tool | Description |
|---|---|
add_table_to_model |
[DESTRUCTIVE] Add Excel table to data model |
refresh_table |
Refresh a single table |
Prompts
| Prompt | Description |
|---|---|
describe_model |
Describe the data model in plain English |
workflow_create_measure |
Guided workflow to create a measure |
workflow_analyze_model |
Analyze model structure and suggest improvements |
workflow_test_dax |
Test DAX before creating permanent measures |
workflow_fix_measure |
Debug and fix a broken measure |
workflow_add_relationship |
Guided workflow to add a relationship |
recover_connection |
Recover from a lost Excel connection |
Resources
| URI | Description |
|---|---|
model://schema |
Current Power Pivot data model schema (JSON) |
model://dirty-state |
Check for unsaved changes |
model://logs |
Recent server diagnostic logs |
model://instructions |
Guidelines for working with Power Pivot |
model://best-practices |
Measure best practices guide |
model://dax-guide |
DAX query guide for Excel |
model://workflows |
Common workflow reference |
Performance Tips
Fast Measure Creation
Use autoFormat: false to skip DAX formatting for faster measure creation (~1.5s savings):
{
"tableName": "Sales",
"measureName": "Quick Measure",
"expression": "SUM(Sales[Amount])",
"autoFormat": false
}
Limitations
Excel Power Pivot Limitations
These features are not available in Excel Power Pivot (unlike Power BI):
| Feature | Excel Power Pivot |
|---|---|
| Calculation Groups | ❌ Not supported |
| Perspectives | ❌ Not supported |
| Row-Level Security (RLS) | ❌ Not supported |
| DEFINE COLUMN in DAX queries | ❌ Not supported |
MCP Server Limitations
These exist in Excel but cannot be managed via this MCP due to COM API restrictions:
| Feature | Status |
|---|---|
| Create/Update/Delete Calculated Columns | ❌ Use Power Pivot window |
| Set Column Descriptions | ❌ Use Power Pivot window |
Troubleshooting
| Error | Solution |
|---|---|
| "Excel is not running" | Open Excel with your workbook |
| "Workbook not found" | Ensure the workbook is open in Excel |
| "No data model" | Create a Power Pivot data model first |
| "Not connected" | Call connect_workbook first |
License
MIT License - see LICENSE file.
Contributing
Contributions welcome! Please open an issue or pull request.
Acknowledgments
- Model Context Protocol by Anthropic
- DAX Formatter by SQLBI
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
E2B
Using MCP to run code via e2b.