Dynamic Excel MCP Server
# Dynamic Excel MCP Server
Dynamic Excel file generation server using Model Context Protocol (MCP). This server allows LLMs to automatically create Excel files with any structure through dynamic JSON schemas.
## ๐ Features
- โ
Generate Excel files from JSON schemas
- โ
**Dual transport modes**: Local (stdio) and Remote (HTTP/SSE)
- โ
**Deploy anywhere**: VPS, Cloud (AWS, GCP, Heroku), Docker
- โ
Multiple sheets support
- โ
Advanced formatting (styling, borders, colors)
- โ
Data validation and conditional formatting
- โ
Formulas and calculations
- โ
Charts support (limited)
- โ
Page setup and printing options
- โ
S3 and local file storage
- โ
Presigned URLs for secure downloads
- โ
Freeze panes, auto-filter
- โ
Merged cells and row grouping
- โ
API key authentication
- โ
CORS support for web clients
## ๐ฆ Installation
```bash
npm install
npm run build
```
## โ๏ธ Configuration
Create a `.env` file (copy from `.env.example`):
**For Local (Stdio) Mode:**
```env
TRANSPORT_MODE=stdio # Local MCP client mode
STORAGE_TYPE=local
DEV_STORAGE_PATH=./temp-files
LOG_LEVEL=info
```
**For Remote (HTTP/SSE) Mode:**
```env
TRANSPORT_MODE=http # Remote server mode
HTTP_PORT=3000
HTTP_HOST=0.0.0.0
ALLOWED_ORIGINS=* # Or specific domains: https://app.example.com
API_KEY=your-secret-api-key # Optional
STORAGE_TYPE=s3 # or 'local'
AWS_ACCESS_KEY_ID=your_key
AWS_SECRET_ACCESS_KEY=your_secret
AWS_REGION=ap-southeast-1
S3_BUCKET=your-bucket
PRESIGNED_URL_EXPIRY=3600
LOG_LEVEL=info
```
## ๐ง Usage
### ๐ฅ๏ธ Local Mode (Stdio) - For Claude Desktop
Add to your Claude Desktop or MCP client configuration:
**For macOS** (`~/Library/Application Support/Claude/claude_desktop_config.json`):
```json
{
"mcpServers": {
"excel-generator": {
"command": "node",
"args": ["/absolute/path/to/excel-mcp-server/build/index.js"],
"env": {
"STORAGE_TYPE": "local",
"DEV_STORAGE_PATH": "./temp-files",
"LOG_LEVEL": "info"
}
}
}
}
```
**For Windows** (`%APPDATA%\Claude\claude_desktop_config.json`):
```json
{
"mcpServers": {
"excel-generator": {
"command": "node",
"args": ["C:\\path\\to\\excel-mcp-server\\build\\index.js"],
"env": {
"STORAGE_TYPE": "local",
"DEV_STORAGE_PATH": "./temp-files",
"LOG_LEVEL": "info"
}
}
}
}
```
### ๐ Remote Mode (HTTP/SSE) - For Web Apps & Remote Access
**Start the server:**
```bash
# Using environment variable
TRANSPORT_MODE=http npm start
# Or using npm script
npm run start:http
# Or with .env file configured for http mode
npm start
```
**Server endpoints:**
```
http://localhost:3000/health - Health check
http://localhost:3000/info - Server information
http://localhost:3000/sse - SSE endpoint for MCP clients
```
**Example client usage:**
See `examples/client-example.ts` for a complete TypeScript client example using the MCP SDK.
```typescript
import { Client } from '@modelcontextprotocol/sdk/client/index.js';
import { SSEClientTransport } from '@modelcontextprotocol/sdk/client/sse.js';
const transport = new SSEClientTransport(
new URL('http://localhost:3000/sse'),
{
headers: { 'X-API-Key': 'your-api-key' } // If API_KEY is set
}
);
const client = new Client({
name: 'excel-client',
version: '1.0.0',
}, { capabilities: {} });
await client.connect(transport);
const result = await client.callTool({
name: 'generate_excel',
arguments: excelSchema
});
```
**Deployment options:**
- ๐ณ **Docker**: See `DEPLOYMENT.md` for Dockerfile and docker-compose examples
- โ๏ธ **Cloud**: Deploy to AWS, GCP, Heroku, etc.
- ๐ง **VPS**: Use PM2, systemd, or other process managers
- ๐ **Production**: Enable API key auth, configure CORS, use HTTPS
๐ **Full deployment guide**: See [DEPLOYMENT.md](./DEPLOYMENT.md)
### Tool: generate_excel
The server provides one tool: `generate_excel`
**Input Schema:**
```json
{
"file_name": "report.xlsx",
"sheets": [
{
"name": "Sheet1",
"columns": [...],
"data": [...],
"formatting": {...}
}
],
"metadata": {...},
"options": {...}
}
```
## ๐ JSON Schema Structure
### Column Configuration
```json
{
"header": "Column Name",
"key": "data_key",
"width": 20,
"type": "currency",
"format": "#,##0โซ",
"style": {
"font": {"bold": true, "size": 12},
"alignment": {"horizontal": "center"},
"fill": {
"type": "pattern",
"pattern": "solid",
"fgColor": {"argb": "FFFF0000"}
}
}
}
```
### Supported Column Types
- `text`: Plain text
- `number`: Numeric values
- `currency`: Currency format
- `percentage`: Percentage format
- `date`: Date format
- `datetime`: Date and time format
- `boolean`: Boolean values
- `formula`: Excel formulas
### Formatting Options
```json
{
"freeze_panes": "A2",
"auto_filter": true,
"conditional_formatting": [
{
"range": "A2:A100",
"type": "cellIs",
"operator": "greaterThan",
"formulae": [0],
"style": {
"fill": {
"type": "pattern",
"pattern": "solid",
"fgColor": {"argb": "FF90EE90"}
}
}
}
],
"totals_row": {
"column_key": "=SUM(A2:A100)"
},
"merged_cells": ["A1:D1"],
"row_heights": {
"1": 30,
"2": 25
}
}
```
## ๐ Examples
### 1. Simple Data Table
See: `examples/01-simple-table.json`
Creates a basic product table with formatting:
- Freeze panes
- Auto-filter
- Currency formatting
### 2. Financial Report
See: `examples/02-financial-report.json`
Advanced report with:
- Report layout with title
- Conditional formatting
- Percentage calculations
- Formula totals
### 3. Employee Database
See: `examples/03-employee-database.json`
Employee management spreadsheet with:
- Multiple column types
- Date formatting
- Currency display
- Auto-filter
### 4. Multi-Sheet Report
See: `examples/04-multi-sheet-report.json`
Comprehensive report with:
- Multiple sheets
- Summary and detail views
- Cross-sheet consistency
## ๐จ Development
```bash
# Run in development mode (with auto-reload)
npm run dev
# Build TypeScript
npm run build
# Start production server
npm start
# Run tests
npm test
# Lint code
npm run lint
```
## ๐งช Testing with MCP Inspector
Test the server using the MCP Inspector:
```bash
npx @modelcontextprotocol/inspector node build/index.js
```
## ๐ฏ Use Cases
1. **Data Export**: Export database queries to formatted Excel files
2. **Financial Reports**: Generate quarterly/annual financial statements
3. **Inventory Management**: Create product catalogs and stock reports
4. **HR Management**: Employee databases and payroll reports
5. **Sales Analytics**: Sales reports with charts and conditional formatting
6. **Project Tracking**: Project status reports with multiple sheets
## ๐๏ธ Architecture
```
src/
โโโ index.ts # MCP Server entry point
โโโ types/
โ โโโ schema.ts # TypeScript types & Zod schemas
โโโ generators/
โ โโโ base-generator.ts # Abstract base class
โ โโโ basic-generator.ts # Simple tables
โ โโโ report-generator.ts # Reports with styling
โโโ formatters/
โ โโโ cell-formatter.ts # Cell formatting
โ โโโ style-formatter.ts # Styling utilities
โ โโโ formula-builder.ts # Formula generation
โโโ storage/
โ โโโ s3-storage.ts # S3 upload handler
โ โโโ local-storage.ts # Local file system
โโโ validators/
โ โโโ schema-validator.ts # JSON schema validation
โโโ utils/
โโโ logger.ts # Logging utility
โโโ error-handler.ts # Error handling
```
## ๐ Security Notes
- For S3 storage, ensure proper IAM permissions
- Use presigned URLs for temporary file access
- Set appropriate expiry times for download links
- Validate all user inputs through Zod schemas
## ๐ License
MIT
## ๐ค Contributing
Contributions are welcome! Please feel free to submit a Pull Request.
## ๐ง Support
For issues and questions, please open an issue on GitHub.
## ๐ Acknowledgments
Built with:
- [Model Context Protocol (MCP)](https://github.com/anthropics/mcp)
- [ExcelJS](https://github.com/exceljs/exceljs)
- [Zod](https://github.com/colinhacks/zod)
- [AWS SDK](https://aws.amazon.com/sdk-for-javascript/)
TDQS
Scored across 1 tool
With only one tool, there is no possibility of confusion or overlap between tools. The single tool 'generate_excel' has a clear and distinct purpose focused on creating Excel files from JSON schemas.
Since there is only one tool, naming consistency is inherently perfect. The tool name 'generate_excel' follows a clear verb_noun pattern, and there are no other tools to create inconsistencies.
A single tool for a server named 'Dynamic Excel MCP Server' feels thin and under-scoped. While the tool is feature-rich, the domain suggests operations like reading, updating, or analyzing Excel files, which are missing. This limits the server's utility for comprehensive Excel interactions.
The server is severely incomplete for its implied domain of dynamic Excel operations. It only supports generation from JSON schemas, lacking essential CRUD operations such as reading existing files, updating data, or performing analyses. This creates significant gaps that will hinder agent workflows involving Excel beyond initial creation.