Moonwatch.rs

🚧 This is an early development version of the software. 🚧

Moonwatch.rs is a privacy-focused digital wellbeing app. Get insights into how you spend your screen time – you choose what data is tracked and where it is stored.

You can run Moonwatch.rs completely self-hosted on your desktop or laptop; aggregating data from multiple machines is also possible via a network drive or any of the "Shared Folder" cloud services (eg. Dropbox, OneDrive, MEGA, etc.).

Currently, Moonwatch.rs consists only of the moonwatcher daemon, which is a background service recording active window at regular intervals and logging it into .jsonl files. More features including analytics and GUI are planned.

The moonwatcher daemon

Supported platforms

• Linux (and other unix-like systems), GNOME, X11
  • dependencies: gnome-screensaver-command, xprintidle, xdotool
  • tested on Ubuntu 22.04 LTS, Ubuntu 24.04 LTS
• Windows
  • no dependencies
  • tested on Windows 10 22H2, Windows 11

Installation

Tested on Ubuntu 24.04 LTS.

• sudo apt install gnome-screensaver xprintidle xdotool
• Clone the repository.
• ./build_linux.py && ./build/moonwatch-rs_0.1.0_Linux-x86-64/install_unix.py
• This will install into ~/.moonwatcher-rs.
  • It sets up a Systemd user service moonwatcher-rs that starts moonwatcher on startup.
  • Events are written to ~/.moonwatcher-rs/logs
  • To customize, edit ~/.moonwatcher-rs/config.json
  • To check up on the daemon, run systemctl --user status moonwatch-rs
  • To reload config, run systemctl --user reload moonwatch-rs

CLI

moonwatcher config.json

JSON configuration

The overall structure is as follows (relative paths are taken to start in the directory where the JSON config is located):

• "main" (object)
  • "output_dir" (string)
    • path to directory where event logs are stored
  • "sample_every_sec" (number)
    • delay between sampling (seconds)
  • "write_every_sec" (number)
    • delay between writing samples to a file (seconds)
  • "path_to_base_config" (string or null)
    • path to another .json configuration file from which "ignore", "anonymize" and "tags" definitions will be read and added to definitions in this config file
    • this is useful for sharing settings across different systems
• "ignore" (object, array or null)
  • one or more WindowEventMatcher objects (see below)
  • events that match will not be recorded at all
• "anonymize" (object, array or null)
  • one or more WindowEventMatcher objects (see below)
  • events that match will be recorded in redacted from
• "tags" (object)
  • "<tag name>" (object, array or null)
    • one or more WindowEventMatcher objects (see below)
    • events that match will get assigned "<tag name>" in output

A WindowEventMatcher definition is an object with at least one of the following keys:

• "window_title" (string)
  • a regular expression (regex::Regex) that is tested against window title
• "process_path" (string)
  • a regular expression (regex::Regex) that is tested against process path

The WindowEventMatcher definition is used to match events – an event must match all predicates defined by given WindowEventMatcher (AND semantics). If you want OR semantics, just define multiple WindowEventMatchers.

Full configuration example:

{
  "main": {
    "output_dir": "./logs",
    "sample_every_sec": 15,
    "write_every_sec": 21600,
    "path_to_base_config": null
  },
  "ignore": [{
    "window_title": "title to ignore"
  }],
  "anonymize": [{
    "window_title": "title to anonymize"
  }],
  "tags": {
    "youtube": [{
        "window_title": "YouTube — Mozilla Firefox$",
        "process_name": "firefox(\\.exe)?$"
      },
      {
        "window_title": "YouTube — Mozilla Firefox$",
        "process_name": "chrome(\\.exe)?$"
      }
    ],
    "pycharm": {
      "process_path": "JetBrains/Toolbox/apps/PyCharm"
    },
    "clion": {
      "process_path": "JetBrains/Toolbox/apps/CLion"
    }
  }
}