PowerPoint Translator MCP Service

A Model Context Protocol (MCP) service that provides PowerPoint translation capabilities using AWS Bedrock models.

Features

• Translate PowerPoint presentations to multiple languages

• Preserve formatting during translation

• Support for AWS Bedrock Nova models

• Intelligent handling of proper nouns, brand names, and special content

• Full MCP protocol compliance with proper initialization and tool handling

• Fallback implementation when MCP library is not available

Supported Languages

• Simplified Chinese (zh-CN)

• Traditional Chinese (zh-TW)

• English (en)

• Japanese (ja)

• Korean (ko)

• French (fr)

• German (de)

• Spanish (es)

Prerequisites

• Python 3.8+

• AWS account with Bedrock access

• AWS credentials configured (via aws configure or environment variables)

Installation

Quick Setup

Virtual Environment Setup (Recommended)

MCP Configuration

Option 1: AWS CLI Configuration (Recommended)

First, configure your AWS credentials:

Then add to your Amazon Q configuration:

Option 2: Direct Credential Configuration

Usage

With Amazon Q

Once configured, you can use natural language commands:

Command Line Usage

Available Tools

1. translate_ppt - Translate a PowerPoint document

   • input_file: Path to the input PowerPoint file (required)

   • target_language: Target language code (default: zh-CN)

   • output_file: Path to save the translated file (optional)

   • translation_method: Translation method, 'nova' or 'claude' (default: nova)

2. list_supported_languages - List all supported target languages

Command Line Arguments

• --mcp: Run in MCP mode

• --translate: Translate a PowerPoint file

• --input-file: Path to the input PowerPoint file

• --target-language: Target language code

• --output-file: Path to save the translated file

• --model-id: Translation model ID (choices: 'amazon.nova-micro-v1:0', 'amazon.nova-lite-v1:0')

• --list-languages: List supported languages

• --install-deps: Install required dependencies

• --use-uv: Use uv package manager instead of pip

• --venv: Create and use a virtual environment with uv

• --venv-path: Path for the virtual environment (default: ./venv)

Testing

Test the MCP server functionality:

Troubleshooting

Common Issues

1. AWS Credentials: Ensure AWS credentials are properly configured via aws configure or environment variables.

2. Missing Dependencies: Run python server.py --install-deps to install required packages.

3. MCP Protocol Errors: The server implements full MCP protocol compliance. If you encounter initialization errors, ensure you're using a compatible MCP client.

4. File Permissions: Ensure the server.py file has execute permissions: chmod +x server.py

Development

Local Development

1. Clone and setup:

   git clone https://github.com/yourusername/mcpppttranslator.git cd mcpppttranslator python server.py --install-deps --use-uv --venv

2. Test the server:

   python test_mcp.py

MCP Protocol Compliance

This server implements the full MCP protocol specification including:

• Proper initialize method handling

• Standard tools/list and tools/call methods

• Correct JSON-RPC 2.0 response formatting

• Error handling and graceful degradation

• Backward compatibility with legacy MCP methods

License

MIT