Calorie Visualizer

A local skill for meal logging and visual nutrition analysis.

Core Flow

• User sends meal text/photo (upstream extracts calorie/protein values or calls food-db lookup).
• add (or add-food) writes to calorie_data.db.
• After write, visual_renderer.py generates a fresh report image.
• CLI prints REPORT_IMAGE: so the chat layer can send the image.

Daily Goal Priority

• config.daily_goal (manual override)
• TDEE derived from USER.md
• If user explicitly refuses profile data: logging only, no repeated prompting
• Fallback default in renderer: 2000 kcal

CLI Commands

# Add a meal with explicit nutrition values
python3 scripts/calorie_visualizer.py add "food name" 500 25 [--photo /path/to/image.jpg]

# Add from local food database (local-first, online fallback optional)
python3 scripts/calorie_visualizer.py add-food "Subway chicken sandwich" --multiplier 1.0
python3 scripts/calorie_visualizer.py add-food "rice" --offline

# Daily summary
python3 scripts/calorie_visualizer.py summary

# Regenerate report image
python3 scripts/calorie_visualizer.py report

# Config
python3 scripts/calorie_visualizer.py config daily_goal 2000
python3 scripts/calorie_visualizer.py config user_refused_profile True

Dependencies

cd skills/calorie-visualizer
python3 -m pip install -r requirements.txt

• Python libs: html2image, Pillow
• Online fallback: optional USDA API (USDA_API_KEY)
• Database: SQLite (built into Python)
• Rendering: requires system Chromium/Chrome (called by html2image)

Storage

• calorie_data.db (local SQLite)
• No automatic external sync