MealSpector
A game that teaches you what's actually in your food.
The problem
Most people have genuinely poor nutritional intuition. Not because they're careless about food, but because the way we perceive food is shaped by colour, texture, and cultural story rather than actual numbers. You'd be surprised how many people think avocado is more caloric than cheddar. (It isn't: cheddar has roughly 400 kcal per 100g, avocado about 160.) That disconnect is what MealSpector is built to close.
How MealSpector works
MealSpector is a free nutrition comparison game. You're shown two foods and asked a single question: which one has more calories? More fat? More sugar? You pick, the answer is revealed, and both foods show their full nutritional breakdown side by side. Ten questions, three lives, a scoring streak if you're on a roll. The comparisons are designed to surface the counterintuitive ones: a granola bar versus a Snickers, foie gras versus peanut butter, dates versus gummy bears. Round after round, the game builds the kind of intuition that stacks up in a way a nutrition label never could.
What makes it worth coming back
Every day at midnight UTC, everyone playing gets the same 10 questions, generated deterministically from the date with no server required. It's the Wordle mechanic applied to food, and it gives people a reason to come back tomorrow. Your streak is stored locally and survives between sessions. The game covers 340 foods across 21 categories, from Czech cuisine to Japanese street food to protein powders and cocktails. It's fully translated into 11 languages, works offline as a Progressive Web App, and every food has a real photograph.
Where it's going
The idea came from a practical frustration: I wanted to search for meals in restaurants near me, or in recipes, filtered by nutritional data. Something that answers 'where do I go tonight if I want 40g of protein' or 'what can I cook that fits my goals'. That search engine doesn't exist yet in a form I'd actually use.
The game is Phase 1, and the current version was built with Claude over a fun couple of hours. It's just the beginning. The broader goal is a food intelligence platform: search any food, meal, or restaurant dish and instantly understand what you're eating, filtered by nutritional constraints, dietary needs, and personal goals. Phase 2 is a food search engine. Phase 3 adds local restaurant integration. Phase 4 is an AI nutrition advisor, practical and evidence-based. The email list from the game is the foundation for all of it.