Grass Pass: a local AI that gets me off the screen and outside

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

What I Built

Grass Pass is a small web app that turns "I have 30 minutes" into a short outdoor plan.

You choose how much time you have, who is going, and how you feel, then type what is near you (a park, a river bank, paddy fields). It gives you one simple activity with a few things to bring, three steps, three small things to look for along the way, a fallback if it rains, and a last line telling you to put your phone in your pocket and go.

I built it for people like me who spend most of the day on a screen and keep saying they'll go outside later. The app is meant to be the shortest part of the experience: you read the plan for a minute and then close the tab.

How it went outside: [Write 2 or 3 honest sentences. What did you type in? What plan did it give you? Did you do it? What did you notice or enjoy, or what didn't work?]

Demo

Watch the demo here: https://drive.google.com/file/d/1oc-JS8zHFYSKIUqPVGMfTGDnTDO86xX1/view?usp=sharing

Code

Karthikpillai77 / grass-pass

# Grass Pass

A tiny local web app that turns "I have 30 minutes" into a short outdoor plan.
Powered by the open-weight Gemma model running locally through Ollama.
No cloud, no API key, works offline once the model is downloaded.

## Run it
1. Install Ollama: https://ollama.com
2. `ollama pull gemma3:4b` (or `gemma3:1b` for a smaller PC)
3. `python app.py`
4. Open http://localhost:8001

Use a different model (PowerShell): `$env:MODEL="gemma3:1b"; python app.py`

View on GitHub

How I Built It

I'm just starting out, so I kept it small on purpose.

• Model: Gemma 3, an open-weight model from Google.
• Local inference: Ollama runs the model on my own computer.
• Backend: one Python file (app.py) using only the standard library. It serves the page and sends a prompt to Ollama.
• Frontend: one plain HTML file with a small form. It remembers what is near you, so you don't retype it every time.

The prompt does the main work. It tells the model how much time you have, who is going, what is nearby and how much energy you have, then asks for one safe, simple activity under 150 words, so the plan stays quick to read.

Why Does Open Innovation Matter?

An app that sends you outside should work when you're outside. Because Gemma runs locally through Ollama, Grass Pass needs no account and no API key, and it works without an internet connection once the model is downloaded. That matters in places with weak signal, which is exactly where a walk often takes you.

What you type about where you live and what is near your home also stays on my own computer instead of going to someone else's server. It costs nothing to run, and I can switch to a smaller or bigger model by changing one setting, which let me match the model to my laptop. With a closed API I would be paying per request and sending my location details to someone else.

One honest limit: a small local model can make mistakes or suggest something that doesn't suit your area, so use common sense and stay safe outdoors.

Prize Categories

Best Use of Gemma: Gemma 3 runs locally through Ollama and is the entire AI core of Grass Pass. It writes every outdoor plan on my own computer.