Gen-Wal 🧠

A personal experiment in ambient computing.

Gen-Wal explores a simple question:

What if the desktop itself could act as a passive environmental cue — instead of another app, notification, or dashboard?

---

What this is NOT

Gen-Wal is not:

• a productivity app
• a motivational tool
• a notification system
• a dashboard or tracker

There is nothing to click. Nothing to dismiss. Nothing to optimize.

---

The Experiment

I spend a large portion of my day in front of my machine. Keeping goals in my head wasn’t enough, but I also didn’t want yet another app demanding attention.

Gen-Wal changes one visual element of the environment once per day.

Quietly.

It uses a profile (a simple text file describing a mindset or focus) and generates a daily background that acts as a passive reference frame — something you see repeatedly without being interrupted.

This is not meant to push behavior. It’s meant to exist.

---

What It Is

Gen-Wal is a small local daemon written in Python that:

• Reads a text-based profile (e.g. Stoic, Builder, Zen)
• Generates short text using a local or remote LLM
• Optionally generates a matching image
• Updates the desktop wallpaper automatically

You can use it, modify it, or ignore it.

---

Privacy & Local Execution

Gen-Wal runs as a local daemon. It has no accounts, no analytics, and no tracking code.

• Default: Uses free remote APIs (Pollinations.ai) for zero-setup generation.
• Local: Can be configured to run 100% locally using Ollama/LocalAI for text and local directories for images.

You control where the data goes.

---

Installation

One-line install for Linux (GNOME / KDE / Unity):

curl -sL https://gen-wal.laptopserver.dev/install | bash

This sets up:

• The genwal CLI (for easy config & overrides)
• A systemd timer (daily update)
• Default config in ~/.gen-wal/

---

Configuration

Everything is configurable. You can control how visible or subtle the output is.

text_position: "bottom_right"   # understated placement
font_size: 45                   # smaller text
image_provider: "pollinations:image"  # or "local_dir"

watermark:
  enabled: true
  position: "bottom_right"
  opacity: 150

profile_path:
  - "profiles/examples/stoic.md"
  - "profiles/examples/deep_work.md"
  - "profiles/examples/builder.md"
  - "profiles/examples/zen.md"

---

Profiles (Reference Frames)

Profiles are not just quote collections — they define mental reference frames.

Included examples:

• Stoic — Restraint, impermanence, control
  (Meditations, Letters from a Stoic)

• Deep Work — Focus, systems, resistance
  (Deep Work, Atomic Habits, War of Art)

• Builder — Craft, simplicity, iteration
  (Hackers & Painters, Unix philosophy)

• Zen — Presence, patience, non-forcing
  (Zen Mind, Tao Te Ching)

---

Create Your Own Profile

Profiles are plain Markdown files.

---
quote_prompt_template: "Max 12 words. Calm, direct, non-preachy."
image_prompt_template: "Muted, minimal, low-contrast background."
---
# Quiet Focus
Presence over pressure.

Point your config to it and that’s it.

---

CLI Usage

genwal run                         # Generate now
genwal run --font-size 40           # Override settings
genwal config                      # Edit config
genwal profile list                # List profiles
genwal profile use <name>          # Switch profile
genwal profile edit <name>         # Create/Edit profile (Auto-switch)
genwal logs                        # View logs

---

Uninstall

To remove everything:

cd ~/.gen-wal && ./uninstall.sh

---

Future Ideas

• Time-bounded focus windows
• Extended context from books/articles
• Daily intent injection
• Community profile sharing

---

Developing

This is a personal project.

git clone https://github.com/nicemit/gen-wal
cd gen-wal
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
python3 main.py

---

License

MIT

If you find this experiment interesting, a GitHub star helps others discover it.