Daily Thucydides

An automated daily reader of Thucydides' History of the Peloponnesian War with AI-enhanced educational content, published to thucydides.caseyjr.org.

Overview

This project transforms the Crawley translation of Thucydides' history into a ~515-day educational journey, with each day featuring:

• A passage of 2500 characters (~350-500 words)
• Modern translation (sophisticated prose)
• Historical context
• Scholarly annotations with links
• Parallel accounts from other ancient historians
• Related passages within Thucydides
• Discussion prompts
• Reading progress tracking

Project Structure

thucydides/
├── src/                      # Source code
│   ├── parser.py            # Text parsing into chunks
│   ├── enricher.py          # AI content generation
│   ├── publisher.py         # Bear Blog API integration
│   └── utils.py             # Helper functions
├── data/
│   ├── raw/                 # Original Gutenberg text
│   ├── processed/           # Parsed chunks (JSON)
│   └── published/           # Publication tracking
├── scripts/
│   ├── setup.sh             # Initial setup script
│   ├── process_text.py      # One-time text processing
│   └── daily_post.py        # Daily posting script
├── .github/workflows/
│   └── daily-post.yml       # GitHub Actions automation
├── requirements.txt         # Python dependencies
├── config.yaml              # Configuration file
└── README.md

Setup Instructions

Prerequisites

• Python 3.9+
• Anthropic API key (Claude Pro)
• Bear Blog API key
• GitHub account (for automation)

Step 1: Clone and Install

git clone https://github.com/miguelito4/thucydides.git
cd thucydides
pip install -r requirements.txt

Step 2: Configure

Create a .env file with your credentials:

ANTHROPIC_API_KEY=your_anthropic_key
BEAR_BLOG_API_KEY=your_bear_blog_key
BEAR_BLOG_ID=your_blog_id

Step 3: Process the Text

Download and process Thucydides' text:

python scripts/process_text.py

This will:

1. Download the Crawley translation from Project Gutenberg
2. Parse it into 2500-character chunks
3. Generate AI-enhanced content for each chunk
4. Save everything to data/processed/

Step 4: Test Publishing

Test a single post:

python scripts/daily_post.py --test

Step 5: Enable Automation

The GitHub Action will automatically post one passage per day at 6:00 AM UTC.

Set repository secrets:

• ANTHROPIC_API_KEY
• BEAR_BLOG_API_KEY
• BEAR_BLOG_ID

How It Works

Text Processing

The parser divides Thucydides into ~515 logical chunks, respecting:

• Natural paragraph breaks
• Book/chapter divisions
• Narrative flow
• 2500-character target length

AI Enhancement

For each passage, Claude generates:

1. Modern Translation: Sophisticated prose matching the Warner translation style
2. Historical Context: What's happening in the narrative
3. Annotations: Links to scholarly analyses and resources
4. Parallel Accounts: Related passages from Herodotus, Plutarch, Xenophon
5. Internal References: Related sections within Thucydides
6. Discussion Prompts: Thoughtful questions for reflection
7. Progress Tracking: Current position in the overall work

Daily Publishing

GitHub Actions runs daily to:

1. Load the next unpublished passage
2. Format it for Bear Blog
3. Publish via Bear Blog API
4. Track publication status

Development

Running Locally

# Process a single chunk for testing
python src/parser.py --test-chunk 10

# Regenerate enrichment for a specific chunk
python src/enricher.py --chunk-id 42 --regenerate

# Publish a specific chunk
python src/publisher.py --chunk-id 100

Configuration

Edit config.yaml to customize:

• Chunk size parameters
• AI prompts and templates
• Publishing schedule
• Blog post formatting

Bear Blog Integration

Posts are formatted as Markdown with:

• Title: "Day X: [Book and Chapter]"
• Publication date scheduled sequentially
• Tags: thucydides, ancient-greece, history
• Custom CSS for styling (if needed)

License

The original text is in the public domain. AI-generated enhancements are released under MIT License.

Acknowledgments

• Crawley translation from Project Gutenberg
• Bear Blog for hosting
• Anthropic Claude for AI enhancements