Skip to content

laz4rd/tshll

Repository files navigation


tshll

████████╗███████╗██╗  ██╗██╗     ██╗
╚══██╔══╝██╔════╝██║  ██║██║     ██║
   ██║   ███████╗███████║██║     ██║
   ██║   ╚════██║██╔══██║██║     ██║
   ██║   ███████║██║  ██║███████╗███████╗
   ╚═╝   ╚══════╝╚═╝  ╚═╝╚══════╝╚══════╝

turtle shell · crawl into the nest · local LLM chat in your terminal


Setup

pip install -r requirements.txt   # textual + ollama
ollama serve                       # start the nest (separate terminal)
ollama pull llama3                 # pull a model if you haven't
python app.py                      # launch tshll

Keybindings

Key Action
Enter Send message / dismiss splash
Ctrl+N New shell (conversation)
Ctrl+M Switch model (the nest)
Ctrl+B Toggle sidebar
Ctrl+L Clear current shell
Ctrl+J Focus input bar
Ctrl+Q Quit
Esc Close modal

Themes

Ctrl+T toggles between two themes:

  • Electron Highlighter — full dark theme, explicit colors
  • Terminal native — inherits your terminal's own colorscheme ($background, $primary, etc.)

If you want to default to terminal native, swap CSS_PATH = THEME_ELECTRON to CSS_PATH = THEME_TERMINAL in app.py.

File structure

tshll/
├── app.py                # application
├── theme_electron.tcss   # electron highlighter theme
├── theme_terminal.tcss   # terminal-native bare theme
├── requirements.txt
└── README.md

Branding notes

  • tshll — turtle shell (drop the vowels, keep the vibe)
  • the nest — what the local model backend is called
  • shells — individual conversations
  • Turtle emoji 🐢 used throughout as the brand mark

About

A terminal-based LLM chat interface powered by local models. Run conversational AI directly in your terminal with Ollama integration, multiple themes, and a charming personality-driven chatbot called "tutel".

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages