05 / 14AI Meal Planning AppLive
MealMaster AI
1 min readmealmasterai.app- Dietary patterns
- 12+ allergies, conditions, budget
- Inputs
- Vision fridge photo → ingredients
- Planner
- LLM constraint engine, hard rules
- Built with
- Codex OpenAI as primary dev loop
Overview
Vision model identifies ingredients from fridge photos; LLM-driven planner reasons over allergies, medical conditions, budget, and equipment before selecting recipes. Constraint engine treats dietary needs as hard logic. 12+ dietary patterns.
How it works
- Vision model reads fridge photos and extracts an ingredient inventory.
- LLM planner reasons over allergies, conditions, budget, and kitchen equipment before selecting meals.
- Constraint engine treats dietary rules as hard logic — not soft prompt suggestions.
- Supports 12+ dietary patterns out of the box.
- Built solo with OpenAI Codex as the primary dev loop.
Vision
LLM
Next.js
OpenAI Codex