Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Excel Analytics MCP Server show me the top 10 customers by revenue from the sales table"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
A self-evolving data analytics toolkit for Claude Desktop. Drop your Excel files in, ask questions in plain English, and build up a personal library of reusable analysis tools — no coding required.
Upload Excel/CSV files via a web dashboard
Ask Claude questions about your data in natural language
Save analyses as reusable tools you can run again and again
Create custom tools that Claude can use on your behalf
Everything stays local — your data never leaves your machine
Related MCP server: csv-mcp-server
Requires Python 3.10+ and uv.
# Install uv if you don't have it
curl -LsSf https://astral.sh/uv/install.sh | sh
# Install and configure Claude Desktop
uvx excel-analytics-mcp --setupThat's it. Restart Claude Desktop and start asking questions.
git clone https://github.com/blakethom8/excel-mcp.git
cd excel-mcp
bash install.shThe install script will:
Check for Python 3.10+ and install uv if needed
Install all dependencies in an isolated virtual environment
Configure Claude Desktop automatically
Create the data directory at ~/Documents/Excel Analytics/
Open the dashboard at http://localhost:8765 (starts automatically when Claude Desktop launches the server) and drag your Excel or CSV files into the upload zone.
In Claude Desktop, try prompts like:
"What datasets do I have?"
"Describe the sales table"
"Show me the top 10 customers by revenue"
"What's the average order value by month?"
"Find all invoices over $5,000 from last week"
When you find a useful query, ask Claude to save it:
"Save this as a tool called 'Top Customers'"
"Create a reusable analysis for monthly revenue trends"
Visit the Tools tab in the dashboard to:
View all your saved analyses and custom tools
Edit tool parameters
Test tools with different inputs
Delete tools you no longer need
Claude Desktop ←→ MCP Server (stdio) ←→ SQLite DB
↓
REST API (port 8765) ←→ Web DashboardMCP Server communicates with Claude Desktop via stdio (the Model Context Protocol)
REST API runs in a background thread on localhost:8765 to power the dashboard
SQLite stores your data locally — Excel/CSV files are converted on upload
Both the MCP server and dashboard share the same database and tool registry
When you upload an Excel file:
Each sheet becomes a separate SQLite table
Column names are cleaned (lowercase, underscores)
Numeric columns are auto-detected and properly typed
Headers are auto-detected (no manual configuration needed)
Tool | Description |
| Show all loaded tables with row/column counts |
| Column names, types, sample values, and basic stats |
| Run read-only SQL queries against your data |
| Statistical summary of a table or specific column |
Tool | Description |
| Save a SQL query as a reusable named tool |
| Build a custom Python tool (sandboxed) |
| See all your saved tools |
| Update an existing tool |
| Remove a tool |
| Run a tool with test parameters |
Tools you create (via Claude or the dashboard) are saved as JSON files and automatically registered as MCP tools on startup. They persist across sessions and can be shared.
The web dashboard at http://localhost:8765 provides:
Data Manager — upload files, browse tables, preview data, see column stats
Tool Workshop — view core tools, manage saved analyses, test tools with auto-generated forms
The dashboard is a single HTML file with no build step — it works in any modern browser.
All data stays on your machine in a visible, browsable folder:
Path | Purpose |
| SQLite database with all your uploaded data |
| Saved analyses and custom tools (JSON files) |
| Generated output files |
| User configuration |
You can override the base directory with the EXCEL_MCP_HOME environment variable.
SQL: Read-only queries only (SELECT). No writes, drops, or schema changes.
Python tools: Sandboxed execution with restricted imports. No file system access, no network calls, no dangerous operations.
Local only: All data stored in ~/Documents/Excel Analytics/. Nothing is sent to external servers. The MCP server communicates only with Claude Desktop on your machine.
No telemetry: No usage tracking, no analytics, no phone-home.
Edit ~/Documents/Excel Analytics/config.json:
{
"port": 8765,
"auto_scan": false
}Setting | Default | Description |
|
| Dashboard port |
|
| Auto-scan a directory for Excel files on startup |
git clone https://github.com/blakethom8/excel-mcp.git
cd excel-mcp
uv sync
uv run python -m excel_mcpPython 3.10+
macOS, Windows, or Linux
MIT
Create a custom Python tool (sandboxed). The code must define a run(db, **kwargs) function.
parameters should be a JSON string like: {"param": {"type": "string", "description": "...", "default": ...}}
Inputs: name (required), description (required), python_code (required), parameters. Returns: result
Remove a user-created tool.
Inputs: tool_id (required). Returns: result
Column names, types, sample values, and basic stats for a table.
Inputs: table (required). Returns: result
Update an existing saved analysis or custom tool.
Inputs: tool_id (required), name, description, sql_template, +2 more. Returns: result
Show all loaded tables with row/column counts.
Inputs: no named parameters. Returns: result
Show all user-created tools (saved analyses and custom tools).
Inputs: no named parameters. Returns: result
Run a read-only SQL query against the database. Only SELECT queries allowed.
Inputs: sql (required). Returns: result
Save a SQL template as a reusable named tool.
parameters should be a JSON string like: {"param": {"type": "string", "description": "...", "default": ...}}
Inputs: name (required), description (required), sql_template (required), parameters. Returns: result
Quick statistical summary of a table or a specific column.
Inputs: table (required), column. Returns: result
Run a tool with given parameters and return results.
params should be a JSON string of the parameters to pass.
Inputs: tool_id (required), params. Returns: result
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 10 tools
The tools have some clear distinctions, but there is notable overlap between create_tool, save_analysis, and edit_tool, which all involve managing custom tools or analyses, potentially causing confusion. Other tools like describe_dataset and summarize serve distinct purposes, but the tool management cluster lacks clear boundaries.
Most tools follow a consistent verb_noun pattern (e.g., create_tool, delete_tool, list_datasets), with only minor deviations like 'query' and 'summarize' being single words. Overall, the naming is readable and predictable, though not perfectly uniform.
With 10 tools, the count is well-scoped for an Excel analytics server, covering data exploration, querying, and custom tool management. Each tool appears to serve a specific role without obvious bloat or redundancy.
The toolset covers core analytics workflows: data inspection (list_datasets, describe_dataset, summarize), querying (query), and custom tool lifecycle (create, edit, delete, test, list). A minor gap is the lack of data manipulation tools (e.g., update or insert), but this is reasonable for a read-focused analytics server.
Interactive data visualization and executive PDF dashboards for ChatGPT, Claude, and Cursor.
AI data analyst: ask about your Excel, CSV, GA4 or database data and get charts and dashboards back
Open, inspect, filter, edit and convert xlsx and csv files from your AI chat. Processing is local.
Open, inspect, filter, edit and convert xlsx and csv files from your AI chat. Processing is local.