This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass.
What I Built
I built TouchGrass AI, a local AI-powered nature mission planner that encourages people to spend less time scrolling and more time outdoors.
Users can choose their preferred outdoor activity, available time (15, 30, or 60 minutes), and difficulty level. The app generates a personalized nature mission using a locally running AI model.
The project includes a Flask backend, a simple web interface, input validation, and browser local storage for tracking self-reported progress.
Demo
- Demo video: https://youtu.be/SX3HJ4AzR5Q
- GitHub repository: https://github.com/utkarshydv27/TouchGrassAI
- Original project article: https://dev.to/ut_yadav_18820e71996/-touchgrass-ai-less-scrolling-more-living-2dpi
Code
The complete source code is available on GitHub:
https://github.com/utkarshydv27/TouchGrassAI
The repository contains the Flask application, frontend files, and automated tests.
How I Built It
I used Python, Flask, HTML, CSS, JavaScript, and Ollama with the Qwen 2.5 1.5B model.
The Flask backend validates user preferences and sends requests to the locally running Ollama service. The frontend displays the generated nature mission and uses browser local storage to track self-reported progress.
I added automated tests covering the home page, invalid inputs, successful mission generation, and Ollama service failures. All 6 tests passed.
Why Does Open Innovation Matter?
Open innovation makes AI development more accessible. By using open-source tools and a locally runnable model, I could build and test an AI-powered application without relying on a paid hosted AI API.
Others can inspect the code, run the project locally, learn from the implementation, and contribute improvements. TouchGrass AI explores how AI can encourage healthier offline habits instead of simply increasing screen time.
Thanks for checking out TouchGrass AI! 🌿
Top comments (0)