# Loop MCP Server
An MCP (Model Context Protocol) server that enables LLMs to process arrays item by item with a specific task.
## Overview
This MCP server provides tools for:
- Initializing an array with a task description
- Fetching items one by one or in batches for processing
- Storing results for each processed item or batch
- Retrieving all results (only after all items are processed)
- Optional result summarization
- Configurable batch size for efficient processing
## Installation
```bash
npm install
```
## Usage
### Running the Server
```bash
npm start
```
### Available Tools
1. **initialize_array** - Set up the array and task
- `array`: The array of items to process
- `task`: Description of what to do with each item
- `batchSize` (optional): Number of items to process in each batch (default: 1)
2. **get_next_item** - Get the next item to process
- Returns: Current item, index, task, and remaining count
3. **get_next_batch** - Get the next batch of items based on batch size
- Returns: Array of items, indices, task, and remaining count
4. **store_result** - Store the result of processing
- `result`: The processing result (single value or array for batch processing)
5. **get_all_results** - Get all results after completion
- `summarize` (optional): Include a summary
- Note: This will error if processing is not complete
6. **reset** - Clear the current processing state
### Example Workflows
#### Single Item Processing
```javascript
// 1. Initialize
await callTool('initialize_array', {
array: [1, 2, 3, 4, 5],
task: 'Square each number'
});
// 2. Process each item
while (true) {
const item = await callTool('get_next_item');
if (item.text === 'All items have been processed.') break;
// Process the item (e.g., square it)
const result = item.value * item.value;
await callTool('store_result', { result });
}
// 3. Get final results
const results = await callTool('get_all_results', { summarize: true });
```
#### Batch Processing
```javascript
// 1. Initialize with batch size
await callTool('initialize_array', {
array: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10],
task: 'Double each number',
batchSize: 3
});
// 2. Process in batches
while (true) {
const batch = await callTool('get_next_batch');
if (batch.text === 'All items have been processed.') break;
// Process the batch
const results = batch.items.map(item => item * 2);
await callTool('store_result', { result: results });
}
// 3. Get final results
const results = await callTool('get_all_results', { summarize: true });
```
### Running the Example
```bash
node example-client.js
```
## Integration with Claude Desktop
Add to your Claude Desktop configuration:
```json
{
"mcpServers": {
"loop-processor": {
"command": "node",
"args": ["/path/to/loop_mcp/server.js"]
}
}
}
```
