parsehealthlog 0.1.6

📓 Transform health journal entries into structured, validated data 🏥

Documentation

Overview

parsehealthlog is a data extraction and curation tool that transforms unstructured health journal entries into structured, validated data ready for downstream analysis.

What it produces:

• health_log.md — All processed entries (newest to oldest) with labs and exams integrated

The tool processes, validates, and enriches health log entries. Reports, summaries, and recommendations are left to downstream consumers of the structured data.

Features

• Parallel processing of hundreds of journal entries
• Lab result integration with automatic interpretation
• Hash-based caching for efficient incremental rebuilds
• Multi-model support via OpenRouter (GPT-4, Claude, etc.)
• Profile-based configuration for managing multiple health logs

Quick Start

# Install uv (if needed)
curl -LsSf https://astral.sh/uv/install.sh | sh

# Clone and install
git clone https://github.com/tsilva/parsehealthlog.git
cd parsehealthlog
uv sync

# Configure
cp .env.example .env
# Edit .env with your OPENROUTER_API_KEY

# Create a profile (profiles/myprofile.yaml)
# health_log_path: /path/to/health.md
# output_path: /path/to/output

# Run
uv run python main.py --profile myprofile

Output Structure

OUTPUT_PATH/
├── health_log.md            # PRIMARY: All entries (newest to oldest)
└── entries/                 # INTERMEDIATE (kept for caching)
    ├── YYYY-MM-DD.raw.md
    ├── YYYY-MM-DD.processed.md
    └── YYYY-MM-DD.labs.md

Configuration

Environment Variables (.env)

Profile Configuration (profiles/<name>.yaml)

See docs/pipeline.md for all configuration options.

Health Log Format

Your health log should be a markdown file with dated sections:

### 2024-01-15

Visited Dr. Smith for annual checkup. Blood pressure 120/80.
Started vitamin D 2000 IU daily.

### 2024-01-20

Feeling better after starting vitamin D. Energy levels improved.

License

MIT