🌿 Touch Grass: An Open-Source AI Nature Companion to Get You Offline and Outdoors

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

🌿 Touch Grass — Your Nature Companion

What I Built

I built Touch Grass, a nature-themed web application designed to encourage people to spend less time in front of screens and more time exploring the real world.

As students and developers, we often spend hours using laptops, smartphones, and other digital devices. Technology helps us learn and create, but spending all our free time online can make us forget the simple things around us, such as taking a walk, observing trees, listening to birds, or enjoying fresh air.

I wanted to build something that uses technology to encourage people to take a break from technology itself.

Touch Grass makes this idea simple and enjoyable through small outdoor challenges and activity suggestions. Instead of asking people to change their entire routine, it encourages them to start with small, achievable actions.

🌱 Key Features

• Nature Challenges: Discover simple outdoor activities, such as taking a short walk, observing a tree, listening to birds, or watching the sunset.
• Mood-Based Suggestions: Choose your mood and available time to receive a suitable activity recommendation.
• Progress Tracking: Complete challenges, earn points, and track your progress.
• Nature Journal: Write down observations and memorable experiences from your time outdoors.
• Local Data Storage: Save supported progress and journal entries in the browser using local storage.
• Responsive Interface: Access the application through a modern web browser on desktop and mobile screens.

🎯 Who Is It For?

Touch Grass is for students, developers, remote workers, and anyone who spends a lot of time indoors or on digital devices and wants an easy way to reconnect with nature.

The main idea is simple: technology should not only keep us connected to our screens; it can also encourage us to step away from them.

The current version is a working front-end prototype. Its activity recommendation system uses predefined JavaScript rules. Integrating an open-source AI model for more personalized recommendations is a planned next step.

Demo

🌐 Live Website: https://aiman00-coder.github.io/Touch-Grass/

💻 GitHub Repository: https://github.com/Aiman00-coder/Touch-Grass

You can visit the live website to explore the interface and try the available features.

Code

The source code is available on GitHub:

🔗 https://github.com/Aiman00-coder/Touch-Grass

The project is built with separate HTML, CSS, and JavaScript files to keep the code easy to understand, maintain, and extend.

Technology Stack

• HTML5: Structures the application and its content.
• CSS3: Creates the visual design, layout, and responsive styling.
• JavaScript: Powers the interactive features, activity recommendations, and progress tracking.
• localStorage: Stores supported user progress and journal entries in the browser.

Project Structure

Touch-Grass/
├── index.html
├── style.css
├── script.js
└── README.md

I kept the implementation lightweight so that it can run directly in a browser without requiring a complicated installation process.

How I Built It

I built Touch Grass using HTML, CSS, and JavaScript, with the goal of making a simple, accessible, and interactive nature companion.

I started by designing the application's structure in HTML, then used CSS to create its nature-inspired appearance and responsive layout. JavaScript handles the interactions, including activity recommendations, challenge completion, points, and journal functionality.

The activity recommendation feature currently uses predefined rules based on the user's selected mood and available time. The application also uses browser local storage to retain supported progress and journal data between visits.

My Open-Source AI Roadmap

The longer-term goal is to make Touch Grass more personalized with an open-source AI model.

One approach I am considering is running a compatible open-weight language model locally through Ollama, with a small backend connecting the model to the web application.

The planned workflow would be:

1. The user selects their mood and the time they have available.
2. The application sends these preferences to a local backend.
3. An open-weight AI model generates a suitable outdoor activity.
4. The application displays the suggestion.
5. The user can try the activity and record the experience in their nature journal.

This would allow recommendations to be more flexible than a fixed list of predefined activities.

Transparency note: The current version does not yet integrate an AI model. The recommendation system is rule-based, and the open-source AI integration is planned future work. I want to develop this further so that open-source AI becomes a real part of the application's functionality rather than simply adding an AI label.

Why Does Open Innovation Matter?

Open innovation matters because it gives developers the freedom to learn, experiment, adapt, and build on existing technology.

As a first-year B.Tech CSE (AI/ML) student, I am still learning how to turn an idea into a working project. Open-source tools make this learning process more accessible by allowing developers to study implementations, understand how systems work, and improve their projects without having to build every component from scratch.

For Touch Grass, open-weight AI models could make personalized recommendations possible without requiring every interaction to depend on a closed, proprietary AI API. Running a model locally could also offer greater control over data handling and reduce dependence on an external inference provider, although hardware requirements and local setup would still need to be considered.

Open-source AI also gives developers the opportunity to experiment with different models and choose an approach that fits their project's needs.

I believe innovation should not be limited to people who already have access to expensive tools or large development teams. Open technologies help beginners like me learn by building real projects, understand the limitations of their work, and improve their ideas step by step.

For me, this challenge is an opportunity to explore how open-source AI can be used to solve a small but relatable problem: helping people spend more time outside.

My Agent Session

I do not have a DevRelay agent session to share for this submission yet.

I may add a session link here if I record one during further development.

Prize Categories

I will update this section according to the eligible partner categories and the technologies actually used in the project.

At present, the project is a front-end prototype built with HTML, CSS, and JavaScript. Open-source AI integration is planned but has not yet been implemented, so I am not claiming an AI partner category based on an integration that is not currently present.

---

Final Thoughts 🌿

Touch Grass started with a simple thought: what if we used technology to remind ourselves to spend a little less time using technology?

I wanted to build something small, useful, and easy to try. This project is also part of my learning journey as a first-year AI/ML student, and I hope to improve it by exploring open-source AI and adding more personalized features.

Sometimes, the best next step is not opening another tab. It is stepping outside.

Build something meaningful. Take a break. Touch grass. 🌱

Thanks for participating in Hacktoberfest Open-Source AI Challenge Week 1!