Akave MCP Server

A Model Control Protocol (MCP) server that allows AI models to interact with Akave's S3-compatible storage.

Installation

Using npm

# Install globally
npm install -g akave-mcp

# Or use with npx
npx akave-mcp

Using pip

pip install akave-mcp

Configuration

Create a .env file with your Akave credentials:

AKAVE_ACCESS_KEY_ID=your_access_key
AKAVE_SECRET_ACCESS_KEY=your_secret_key
AKAVE_ENDPOINT_URL=your_endpoint_url

Usage

Command Line

# Start the server with default settings
akave-mcp

# Specify port and host
akave-mcp --port 8080 --host localhost

Programmatic Usage

import AkaveMCPServer from 'akave-mcp';

const server = new AkaveMCPServer();

// Start the server
await server.start({
  port: 8000,
  host: '0.0.0.0',
  env: {
    AKAVE_ACCESS_KEY_ID: 'your_access_key',
    AKAVE_SECRET_ACCESS_KEY: 'your_secret_key',
    AKAVE_ENDPOINT_URL: 'your_endpoint_url'
  }
});

// Stop the server
server.stop();

Features

• List buckets in Akave storage
• Read file contents
• Write new files
• Update existing files
• Analyze storage contents

API

Resources

• akave://buckets - List all buckets
• akave://buckets/{bucket_name} - Get bucket info
• akave://buckets/{bucket_name}/files - List files in bucket
• akave://buckets/{bucket_name}/files/{file_key}/content - Get file content
• akave://buckets/{bucket_name}/files/{file_key}/info - Get file metadata

Tools

• analyze_bucket(bucket_name) - Analyze bucket contents
• analyze_file_content(bucket_name, file_key) - Analyze file contents
• create_bucket(bucket_name) - Create new bucket
• write_file(bucket_name, file_key, content) - Write file
• append_to_file(bucket_name, file_key, content) - Append to file
• update_csv_row(bucket_name, file_key, row_data) - Update CSV file

Development

# Install dependencies
npm install

# Build TypeScript
npm run build

# Run in development mode
npm run dev

License

MIT