DEV Community

Swagatam Das
Swagatam Das

Posted on

Nature Quest 🌿

#devchallenge #hf26challenge #gemma #ai

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass

What I Built

Nature Quest is a small website that gives you something to do outside. You pick how much time you have, and a local open-weight AI model (Gemma, running through Ollama) writes three short nature challenges for you to go and find.

Code

GitHub logo SwagD15 / NatureQuest

Nature Quest is a small website that gives you something to do outside. You pick how much time you have, and a local open-weight AI model (Gemma, running through Ollama) writes three short nature challenges for you to go and find.

Nature Quest 🌿

Put the screen down. Go touch grass.

Nature Quest is a small website that gives you something to do outside. You pick how much time you have, and a local open-weight AI model (Gemma, running through Ollama) writes three short nature challenges for you to go and find.

Nature Quest screenshot

The Problem

It's easy to say "I should go outside" and then spend the break scrolling instead. A walk is more fun when you have something to look for, but it's hard to come up with ideas on the spot.

The Solution

  1. Choose how much time you have: 5, 10 or 20 minutes.
  2. Click Generate Nature Quest.
  3. Get three simple outdoor challenges, each with a short title and one sentence telling you what to find or notice.
  4. Put your phone away and go find them.

The challenges are written by the AI model each time. They are not…

View on GitHub

How I Built It

i used Gemma 3:270m open-source AI model (open-weight model) through Ollama; first the parson choose there preference (like time - 5 , 10 ,20 minutes , etc.) and app.js reads three choices and builds a prompt, It sends the prompt to Ollama's local API (http://localhost:11434/api/generate) than the Gemma generate the three response in JSON format, The page parses the JSON and shows three game cards.

Why Does Open Innovation Matter?

Nature Quest uses an open-weight model that runs entirely on the user's own machine. There are no API keys, no usage fees and no data sent to a third-party service. Anyone can swap in a different model, change the prompt, or edit the fallback challenges

i am very new to open-weight models, i am learning about the ollama and open-weight models through this projects 😀

Top comments (0)