Track pantry items in real household units and add them by typing, barcode scan, or receipt photo.
Generate a weekly meal plan and shopping list based on what you already have and what is expiring.
Import recipes from a web link deterministically, or from a cooking video using optional AI.
Share one kitchen with your household, giving each family member their own login.
| shockistaken/shuna | 0marildo/imago | 0xdfi/glm-5.2-1m-4x-dgx-spark | |
|---|---|---|---|
| Stars | 3 | 3 | 3 |
| Language | Python | Python | Python |
| Setup difficulty | moderate | easy | hard |
| Complexity | 3/5 | 2/5 | 5/5 |
| Audience | general | general | researcher |
Figures from each repo's GitHub metadata at analysis time.
Requires Docker and Compose, AI, grocery pricing, and Home Assistant integrations are optional extras.
Shuna is a self hosted application meant to run your entire household kitchen from one place: what food you currently have, what you plan to cook, the shopping list that follows from that plan, and the recipes themselves. Because it runs on your own hardware rather than a company's servers, your kitchen data stays on your own network unless you deliberately choose to connect an outside AI service. Most existing kitchen apps either live in the cloud and keep your data, or only handle one piece of the puzzle, inventory, planning, or recipes, without connecting them together. Shuna's approach is to combine all four into a single self hosted app, and to make sure every AI powered feature has a plain, deterministic fallback, so the whole app still works fully even with AI turned off completely. Its inventory tracking uses real household units, like a bulb of garlic or a dozen eggs, rather than only weights, and items can be added by typing, scanning a barcode, or photographing a receipt. A weekly meal planner suggests meals based on what you actually have and what is close to expiring, and it can generate a shopping list that only includes what you are short on, price it against your local grocery store, and export it to other tools like Home Assistant. Recipes can be imported from a web link without any AI involved, or from a cooking video using AI assistance. The app also tracks expiring food, predicts when staple items need restocking based on your own past purchases, and supports multiple people sharing one household account, each with their own login. Running Shuna requires Docker and its Compose tool, optional extras include a locally run AI model through a tool called Ollama, a grocery pricing service, and Home Assistant integration, none of which are required for the core app to work. Setup is a matter of copying an example environment file, filling in a few passwords and secrets, and running one Docker command, after which the first account created automatically becomes the administrator. The backend is built with FastAPI and PostgreSQL, the frontend is a installable web app built with SvelteKit, and the whole project is licensed under the GNU Affero General Public License version 3, meaning modified versions run as a public service must also share their source code.
Shuna is a self-hosted app that combines pantry tracking, meal planning, shopping lists, and recipes into one kitchen system with optional local AI.
Mainly Python. The stack also includes Python, FastAPI, PostgreSQL.
AGPLv3: you can use, modify, and self-host this freely, but if you run a modified version as a public service, you must share your source code with its users.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly general.
This repo across BitVibe Labs
Verify against the repo before relying on details.