Commerce Content Generator

This project generates branded content articles based on user-specified parameters. It supports various article types, languages, and advertiser products using Anthropic's Claude AI models.

Live Demo Here

Features

• Generate listicles, product reviews, gift guides, deal roundups, and how-to articles
• Customize voice, length, language, and advertiser products
• Enrich generated copy with product mentions and link placements
• Output copy for review, editing, and commenting
• Convert final copy to HTML with affiliate links
• Utilize predefined writing styles or generate a new style summary from a provided URL
• Get a calendar of local events for a given country and use them in your article generation

Requirements

• Python 3.12+
• UV package manager
• Anthropic API key

Installation

1. Clone the repository
2. Install dependencies using UV:

    uv sync && uv venv

Usage

1. Run the Streamlit app:

   uv run streamlit run comm_con_gen/streamlit_app.py

2. Enter your Anthropic API key in the sidebar configuration panel

3. Select content generation options:

   • Partner ID
   • Publishing Date
   • Article Type (Gift Guide, Product Review, etc.)
   • Audience Country
   • Article Max Word Length
   • Product Category
   • Language
   • LLM Model (see below)
   • Writing Style

Available LLM Models

The application supports the following Anthropic Claude models:

Model Name	Model ID
Claude 3.5 Sonnet	claude-3-5-sonnet-20241022
Claude 3.5 Haiku	claude-3-5-haiku-20241022
Claude 3 Opus	claude-3-opus-20240229
Claude 3 Sonnet	claude-3-sonnet-20240229
Claude 3 Haiku	claude-3-haiku-20240307

Project Structure

comm_con_gen/
├── __init__.py
├── streamlit_app.py    # Main Streamlit application
├── utility.py          # Core functionality and API interactions
├── countries.py        # Country-specific language mappings
├── guides/            # Article format templates
│   ├── giftguide.txt
│   ├── howto.txt
│   ├── listicle.txt
│   ├── product_review.txt
│   └── dealradar.txt
├── prompts/           # LLM prompt templates
│   ├── country_events.txt
│   └── get_voice_prompt.txt
├── test_products/     # Sample product data
│   └── *.json
├── outputs/          # Generated markdown articles
├── html_outputs/     # Generated HTML articles
└── voices/          # Cached writing style analyses

Key Components

• streamlit_app.py: Main application interface with API key management and content generation options
• utility.py: Core functionality including:
  • LLM model management and API interactions
  • Article generation and formatting
  • HTML conversion and affiliate link processing
  • Publishing calendar generation
  • Writing style analysis
• countries.py: Manages country-specific language mappings and localization
• guides/: Contains templates for different article types
• test_products/: Sample product data in JSON format with details like name, description, price, etc.

Development

The project uses:

• UV for dependency management
• pyproject.toml for project configuration
• Type hints throughout the codebase
• Google-style docstrings
• Async/await for API operations

TODO

• Add support for image formatting and placement
• Expand article type templates
• Add more language support
• Implement caching for API responses
• Add unit tests and integration tests