Large File MCP ServerIntelligent Large File Handling

Quick Example ​

// Read a specific chunk of a large file
{
  "tool": "read_large_file_chunk",
  "arguments": {
    "filePath": "/var/log/system.log",
    "chunkIndex": 0,
    "includeLineNumbers": true
  }
}

// Search for patterns with context
{
  "tool": "search_in_large_file",
  "arguments": {
    "filePath": "/var/log/error.log",
    "pattern": "ERROR.*database",
    "regex": true,
    "contextBefore": 3,
    "contextAfter": 3
  }
}

// Navigate to specific line
{
  "tool": "navigate_to_line",
  "arguments": {
    "filePath": "/code/app.ts",
    "lineNumber": 1234,
    "contextLines": 10
  }
}

Why Large File MCP Server? ​

Working with large files in AI applications presents unique challenges:

• Memory Constraints: Loading entire files into memory is impractical for large datasets
• Context Windows: LLMs have limited context windows that can't accommodate entire files
• Navigation: Finding specific information in massive files requires intelligent search
• Performance: Processing must be fast enough for interactive AI applications

Large File MCP Server solves these problems with:

✅ Smart Chunking - Automatically optimizes chunk size based on file type ✅ Streaming Architecture - Process files of any size without memory issues ✅ Intelligent Caching - LRU cache with 80-90% hit rates for repeated access ✅ Powerful Search - Regex support with contextual results ✅ Type Safety - Full TypeScript support prevents runtime errors

Performance Benchmarks ​

Supported File Types ​

The server intelligently detects and optimizes for:

• Text files (.txt) - 500 lines/chunk
• Log files (.log) - 500 lines/chunk
• Code files (.ts, .js, .py, .java, etc.) - 300 lines/chunk
• CSV files (.csv) - 1000 lines/chunk
• JSON files (.json) - 100 lines/chunk
• XML files (.xml) - 200 lines/chunk
• Markdown files (.md) - 500 lines/chunk
• Config files (.yml, .yaml, .sh) - 300 lines/chunk

Get Started ​

Install via npm:

npm install -g @willianpinho/large-file-mcp

Or use with npx:

npx @willianpinho/large-file-mcp

Community ​

• GitHub Repository
• Issue Tracker
• Contributing Guide

License ​

MIT