Agentic Website Manager - Quick Start Guide

🚀 Quick Setup

1. Navigate to the directory:

   cd scripts/agentic_website_manager
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Set up your API key:

   export GEMINI_API_KEY="your-gemini-api-key-here"
   

4. Test the system:

   python test_system.py
   

5. Run the system:

   # Full update (publications + research + portfolio)
   python run.py --full-update
   
   # Research and portfolio only (preserve existing publications)
   python run.py --research-only
   
   # Skip validation for faster execution
   python run.py --full-update --no-validation
   

🎯 What This System Does

The Agentic Website Manager is a modular replacement for the original website_content_manager.py that:

• 🕷️ WebAgent: Searches arXiv for your papers and downloads PDFs
• 🔍 ParseAgent: Extracts figures and text from research papers
• 📝 ContentAgent: Classifies papers into categories and generates summaries
• 🔍 CriticAgent: Validates all work for quality and consistency
• 🎯 Orchestrator: Coordinates everything in smooth workflows

⚙️ Customization

Edit Search Terms

In config.py, modify:

config.search_config['search_terms'] = [
    'au:"Your Name"',
    'au:"Name, Your"'
]

Customize Research Categories

In config.py, modify:

config.classification_config['categories']['new-category'] = {
    'name': 'New Research Area',
    'description': 'Description',
    'keywords': ['keyword1', 'keyword2']
}

Edit LLM Prompts

In config.py, modify:

config.prompts['paper_classification'] = """
Your custom prompt here...
"""

🔧 Troubleshooting

Common Issues:

1. Import errors: Make sure you’re in the right directory and dependencies are installed
2. API key error: Ensure GEMINI_API_KEY environment variable is set
3. Permission errors: Check file system permissions for writing to directories

Debug Mode:

import logging
logging.basicConfig(level=logging.DEBUG)

📊 System Architecture

WebAgent → ParseAgent → ContentAgent → CriticAgent
    ↓         ↓           ↓             ↓
  Papers → Figures → Categories → Validation

All coordinated by the Orchestrator which handles:

• Pipeline execution
• Error handling
• Progress tracking
• File writing

🎉 Success!

If the test passes, your system is ready to automatically:

• Find and download your research papers
• Extract scientific figures
• Classify papers into research areas
• Generate beautiful research page content
• Create detailed portfolio pages
• Validate everything for quality

Enjoy your automated website management! 🚀