TouchGrass AI: Less Scrolling, More Living 🌿

Описание и применение

TouchGrass AI – локальное веб‑приложение, генерирующее персонализированные задания на свежем воздухе, чтобы сократить время, проведённое в интернете.

Задача

Users spend too much time scrolling on screens instead of engaging in outdoor activities.

Результат

All six automated tests passed, confirming mission generation, input validation, and error handling work as intended.

Рабочий процесс

• 1. User opens the local web app.
• 2. User selects desired mission duration (15, 30, or 60 minutes).
• 3. User inputs interests and experience level via form.
• 4. Flask backend validates input and sends a prompt to the local Ollama service.
• 5. Ollama runs the Qwen 2.5 1.5B model and returns a mission description.
• 6. Frontend displays the mission and stores progress in browser local storage.
• 7. User follows the mission outdoors and can track completion locally.

Доказательства

TouchGrass AI helps people choose an outdoor activity based on their available time, interests, and experience level.

Источник

I built the application using Python, Flask, HTML, CSS, JavaScript, Ollama, and the Qwen 2.5 1.5B model.

Источник

All six automated tests passed.

Источник

Mission-generation prompts are sent to a locally running Ollama service rather than a hosted AI API.

Источник

Once everything is installed, mission generation can work without internet access while the local model service is available.

Источник