Skip to content

Latest commit

 

History

History
382 lines (255 loc) · 8.81 KB

README.md

File metadata and controls

382 lines (255 loc) · 8.81 KB
Raw

Zsh Ollama Command Helper

License: MIT GitHub stars GitHub issues GitHub forks Contributors Last Commit Python PRs Welcome

Stargazers

This project enhances your Zsh terminal by allowing you to input natural language queries for shell commands you can't remember. By pressing Ctrl+B, your query is sent to an Ollama model, which generates the appropriate command. The command is displayed, and you're prompted to execute it or not (y/n).

🎮 Demo

demo.mp4

💡 Simply type your question and press Ctrl+B to get the command you need!

Table of Contents

✨ Features

🗣️ Natural Language Queries

Ask questions in plain English about shell commands

🤖 Model Integration

Uses your locally running Ollama instance with a finetuned model

⚡ Command Execution

Shows the generated command and prompts you to execute it

🔄 Model Selection

Easily switch between different Ollama models

🎨 Customizable

Adjust colors and default settings to your preference

🔒 Privacy-Focused

100% local execution - your queries never leave your machine

Prerequisites

  • Operating System: Unix-like system (Linux, macOS).

  • Shell: Zsh.

  • Python: Version 3.11.

  • Ollama: Installed and running locally.

  • jq: Command-line JSON processor.

Installation

1. Ensure Prerequisites Are Met

Install Zsh

If you don't have Zsh installed:

# On Ubuntu/Debian
sudo apt update
sudo apt install zsh

# On macOS (using Homebrew)
brew install zsh

Install Python 3.11

# On Ubuntu/Debian
sudo apt update
sudo apt install python3.11 python3.11-venv

# On macOS (using Homebrew)
brew install [email protected]

Install Ollama

Follow the installation instructions from Ollama's official documentation.

Install jq

# On Ubuntu/Debian
sudo apt update
sudo apt install jq

# On macOS (using Homebrew)
brew install jq

2. Run the install.sh Script

./install.sh

3. Reload Your Zsh Configuration

source ~/.zshrc

Configuration

Default Ollama Model

The default model is set to vitali87/shell-commands. You can change it by editing the export ZSH_OLLAMA_MODEL line in your ~/.zshrc.

export ZSH_OLLAMA_MODEL="your-preferred-model"

Usage

Ask a Question

Type your natural language query directly into the terminal prompt.

how to list all files modified in the last 24 hours

Trigger the Helper

Press Ctrl+B to activate the helper.

Example Output:

🤔 Asking Ollama (using model: vitali87/shell-commands)... 
Your query: how to list all files modified in the last 24 hours 
Generated command: find . -type f -mtime -1 
Execute? [y/N]

Execute the Command

Press y and hit Enter to execute the command. Press any other key to abort.

Customization

Changing the Model

List Available Models

set_ollama_model

Set a Different Model

set_ollama_model your_model_name
Example:
set_ollama_model llama2:7b

Customizing Colors

The colors can be adjusted by modifying the color codes in the Zsh configuration.

Color Codes:

  • Black: \e[30m
  • Red: \e[31m
  • Green: \e[32m
  • Yellow: \e[33m
  • Blue: \e[34m
  • Magenta: \e[35m
  • Cyan: \e[36m
  • White: \e[37m
  • Reset: \e[0m

Steps:

Open the Zsh configuration file:

nano ~/.zshrc

Locate the ollama_command_helper function.

Modify the echo statements:

echo -e "\e[33mYour query:\e[0m $user_query" 
echo -e "\e[32mGenerated command:\e[0m $command"

Replace the color codes with your preferred ones.

Save and exit the editor.

Reload your Zsh configuration:

source ~/.zshrc

Troubleshooting

Error: No Command Generated

Ensure your Ollama server is running and the specified model is available.

Dependencies Not Found

Make sure Python 3.11 and jq are installed on your system.

Virtual Environment Issues

If you encounter issues with the virtual environment:

Remove the existing virtual environment:

rm -rf ~/.config/zsh/ollama_env

Re-run the installation script:

./install.sh

Reload your Zsh configuration:

source ~/.zshrc

Powerlevel10k Warning

If you see a warning related to Powerlevel10k's instant prompt:

Place the Ollama Command Helper configuration after Powerlevel10k initialization in your ~/.zshrc.

Alternatively, disable the instant prompt feature in Powerlevel10k.

📋 Roadmap & TODO

Progress

🚀 Upcoming Features

User Experience

  • 🎨 Add color themes support
    • Dark mode
    • Light mode
    • Terminal-native theme
  • ⌨️ Customizable keyboard shortcuts
  • 💾 Command history with search functionality
  • 🔍 Auto-completion suggestions

AI/ML Enhancements

  • 🧠 Context-aware command suggestions
  • 📊 Learning from user corrections

Performance & Integration

  • ⚡ Improve response time
  • 🔌 Plugin system for extensions
  • 📦 Package for different package managers
    • Homebrew
    • apt
    • pip

🔄 In Progress

User Feedback System

  • Basic feedback collection
  • 👍 Command rating system (thumbs up/down)
  • 📝 Feedback submission UI
  • 📊 Analytics dashboard for feedback

Command History Enhancement

  • Basic history storage
  • 🔍 Searchable command history
  • 📈 Usage statistics
  • 🎯 Success/failure tracking

🎯 Future Goals

Community Features

  • 👥 Command sharing platform
  • 🌟 Popular commands repository
  • 🤝 Community contributions system

Documentation

  • 📚 API documentation
  • 🎥 Video tutorials
  • 👩‍💻 Developer guide
  • 🌍 Internationalization

✅ Completed

  • Basic command generation
  • Model selection interface
  • Installation script
  • Basic error handling

Acknowledgments

Ollama for providing the LLM serving platform.

OpenAI for the openai Python package.

Feel free to contribute to this project by submitting issues or pull requests.

📊 Project Stats

Repo Card Profile Details

👥 Contributors

🛠️ Built With

Python Shell Zsh

🌟 Show your support

Give a ⭐️ if this project helped you!

visitors