CLI-Message

A terminal-based messaging application written in Python. Chat in real time with friends and communities, all from your command line, through a TUI made with textual.

Inspired by Discord ofc lol: servers, channels, DMs and all...

This project mainly served as a learning experience with Textual TUI framework and a bit for SQLite-based databasesss

Table of Contents

• Features
• Requirements
• Installation
• Directory Tree
• Getting Started
• Settings
• AI Usage
• License

Features

• Authentication: Register and login with username/password.
• Servers: Create or join servers using invite codes. Each server has its own channels and member list.
• Channels: The Owner can create text channels inside servers.
• Direct Messages: Send and receive DMs with friends. Includes a friend request system with pending/accepted states.
• Notifications and unread badges: Unread message counts are tracked per channel, per server and per dm + mention notifications.
• Message Features: Reply to messages, edit and delete your own messages, and attach files. Mentions (@username) are highlighted inline and the mentioned user is notified.
• Presence: Set your status to online, dnd, invisible (and offline). Presence is updated automatically on launch and exit.
• Profile Customization: Set a display name, bio, pronouns, status, name color, accent (banner) color and connections.
• Compact Mode: Toggle a condensed message view that groups consecutive messages from the same user.
• Themes: Switch between dark and light UI themes.
• Real-time Updates: Messages, notifications, member lists, and unread badges refresh automatically via polling intervals.
• Image Support: Send and display image attachments inside the chats (uses textual-image, so for some terminals it may not work).
• Settings: Configure user, notifications, theme, and compact mode.
• SQLite Database: All data is stored in a local SQLite3 database.
• Textual TUI: The entire interface is built with the Textual framework waaaaaa

Requirements

• Python: 3.8 or higher
• Dependencies: Install via requirements.txt
  • textual
  • textual-image
  • Pillow
  • rich

Installation

1. Clone the repository and navigate to the project folder:

   git clone https://github.com/artur3333/CLI-Message.git
   cd CLI-Message

2. Install the required Python packages:

   pip install -r requirements.txt

3. Run the app:

   python main.py

The SQLite database (cli_message.db) is created automatically on first run.

Directory Tree

CLI-Message/
├── main.py          # Main application entry point launching the Textual app
├── auth.py          # Authentication & session management
├── db.py            # SQLite database layer
├── screens.py       # Textual screens and UI widgets
├── utils.py         # Utility functions & helpers
├── style.tcss       # Textual CSS stylesheet
└── cli_message.db   # SQLite database (auto-generated on first run)

Getting Started

1. Start the application:

   python main.py

2. Register a new account on the login screen by entering a username and password and clicking Register.

3. DMs tab loads by default browse your friends list, send friend requests, and open conversations (you won't be able to do much because your account is new).

4. Servers can be accessed from the left sidebar. Click + to create a new server or join one with an invite code.

You can open other instances of the app in separate terminal windows and create multiple accounts to test the messaging features between them.

5. Press Ctrl+Q to quit.

Settings

Each user has their own settings, editable from the settings button (S) in the bottom-left corner (near the username).

Setting	Description	Default
dm_notifications	Enable notifications for new DMs	on
mention_notifications	Enable notifications for @mentions	on
compact_mode	Compact grouped messages from the same user	off
theme	UI color theme (dark or light)	dark

AI Usage

Honestly, this project was a pain. Between Textual being unintuitive to structure at first, CSS constantly misbehaving, async refresh bugs, and all of this happening alongside other stuff, AI was consulted more than I originally planned. Most of the core logic was written independently, but there were moments where I was just too stuck to keep going without help. Here is where AI was used:

• Live message updates (for last devlog) Getting the live chat refresh to work without either showing users the wrong messages or causing the entire chat to visually reload mid conversation was a challenge. Full reloads looked awful while someone was actively chatting, but partial updates kept breaking in different ways. AI was used to come up with a solution that would only update the messages that actually changed and keep the rest of the chat visually stable.

• Light theme stylesheet Writing the full light mode style would have taken a long time at a point in the project where I was already trying to wrap things up. AI helped produce the initial version which I then adjusted when was adding new UI elements (after that update).

• Styling and CSS Textual's CSS is a pain, at least for me, and small visual changes sometimes took way longer than expected. AI was used occasionally to fix layout bugs and get things looking fine especially during the servers/channels UI work and the profile panels.

• Bug fixing throughout Various bugs were brought to AI when nothing was working and I wasn't sure how to proceed, it can be wrong message ordering from the database, Textual widget state issues, Terminal rendering problems, and other things that showed up everywhere in the code.

• Documentation This README was partially generated by AI and then edited for accuracy.

License

This project is licensed under the MIT License. See the LICENSE file for details