wilketob/amztec-mcp-server
If you are the rightful owner of amztec-mcp-server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.
A Model Context Protocol (MCP) server that provides seamless access to Amazon's Selling Partner API, enabling AI assistants and applications to retrieve product information, pricing data, and generate listing optimization suggestions for Amazon products.
Amazon SP API MCP Server
A Model Context Protocol (MCP) server that provides seamless access to Amazon's Selling Partner API. This server enables AI assistants and applications to retrieve product information, pricing data, and generate listing optimization suggestions for Amazon products.
Features
- Product Information Retrieval: Get comprehensive product details by SKU including title, description, features, images, attributes, dimensions, and sales rank
- Pricing Data Access: Retrieve current competitive pricing information for products
- Listing Optimization: Analyze product data and provide structured suggestions for improving Amazon listings
- Multi-user Support: Configurable for different seller accounts and user scenarios
- German Marketplace: Pre-configured for Amazon Germany (Marketplaces.DE)
Quick Start
Prerequisites
- Python 3.10 or higher
- Amazon SP API credentials (refresh token, LWA app credentials, AWS credentials, role ARN)
- Valid Amazon seller account with API access
Installation
- Clone the repository:
git clone <repository-url>
cd amztec-mcp-server
- Install dependencies using uv (recommended):
uv sync
Or with pip:
pip install -r requirements.txt
- Set up environment variables:
cp .env.example .env
Then edit .env
with your Amazon SP API credentials:
AMAZON_REFRESH_TOKEN=your_refresh_token
AMAZON_LWA_APP_ID=your_lwa_app_id
AMAZON_LWA_CLIENT_SECRET=your_lwa_client_secret
AMAZON_AWS_ACCESS_KEY=your_aws_access_key
AMAZON_AWS_SECRET_KEY=your_aws_secret_key
AMAZON_ROLE_ARN=your_role_arn
SELLER_ID=your_seller_id
Running the Server
Start the MCP server:
python amazon_mcp_server.py
Test the Amazon SP API connection:
python test.py
MCP Tools
The server exposes three main tools through the MCP interface:
1. get_amazon_product_info
Retrieves comprehensive product information by SKU.
Parameters:
sku
(required): Stock Keeping Unit of the productseller_id
(optional): Seller ID for multi-user scenarios
Returns: JSON object with product title, description, features, images, attributes, dimensions, and sales rank data.
2. get_amazon_product_pricing
Gets current pricing information for a product.
Parameters:
sku
(required): Stock Keeping Unit of the productseller_id
(optional): Seller ID for multi-user scenarios
Returns: JSON object with competitive pricing data from Amazon.
3. optimize_product_listing
Analyzes product data and provides optimization suggestions.
Parameters:
asin
(required): Amazon Standard Identification Numberoptimization_focus
(optional): Focus area - 'title', 'description', 'features', or 'all'user_id
(optional): User ID for multi-user scenarios
Returns: Structured data for external analysis and optimization recommendations.
Configuration
Amazon SP API Setup
- Register as an Amazon SP API developer
- Create a Login with Amazon (LWA) application
- Set up AWS IAM role with necessary permissions
- Obtain refresh token through the authorization workflow
- Configure environment variables in
.env
file
Marketplace Configuration
The server is configured for German marketplace by default. To change the marketplace, modify the Marketplaces
parameter in the AmazonSPAPIClient
initialization:
amazon_client = AmazonSPAPIClient(marketplace=Marketplaces.US) # For US marketplace
Architecture
- Single-file MCP server: All functionality contained in
amazon_mcp_server.py
- AmazonSPAPIClient class: Wrapper for Amazon SP API operations with error handling
- MCP tool definitions: Standardized interface for AI assistant integration
- Async support: Built on asyncio for efficient concurrent operations
Development
Project Structure
amztec-mcp-server/
�� amazon_mcp_server.py # Main MCP server implementation
�� test.py # Test/debug script
�� hello.py # Simple hello world script
�� pyproject.toml # Project configuration
�� .env # Environment variables (not tracked)
�� README.md # This file
Dependencies
mcp[cli]>=1.9.3
- Model Context Protocol frameworkpython-amazon-sp-api>=1.9.36
- Amazon SP API client librarypython-dotenv>=1.1.0
- Environment variable management
Testing
Run the test script to verify Amazon SP API connectivity:
python test.py
Multi-user Support
The server supports multi-user scenarios where different users can have their own Amazon SP API credentials. This is useful for SaaS applications or services managing multiple seller accounts.
Error Handling
The server includes comprehensive error handling for:
- Invalid or missing credentials
- Network connectivity issues
- Amazon API rate limiting
- Product not found scenarios
- Invalid SKU/ASIN formats
License
[Add your license information here]
Contributing
[Add contribution guidelines here]
Support
For issues and questions:
- Check the Amazon SP API documentation
- Review environment variable configuration
- Ensure proper API permissions and rate limits