vity 0.2.5

AI-powered terminal assistant for generating shell commands

🤖 Vity - AI Terminal Assistant

AI-powered terminal assistant that generates shell commands and provides coding help. Works with OpenAI, Google Gemini, local models (Ollama), and any OpenAI-compatible API.

✨ Features

• 🎯 Smart Command Generation: Describe tasks, get exact commands
• 🤖 Multi-Provider Support: OpenAI, Google Gemini, Ollama, or any OpenAI-compatible API
• 🏠 Local Model Support: Run completely offline with Ollama
• 🧠 Context Awareness: Record terminal sessions for better responses
• 💬 Chat Mode: Ask questions about errors and commands
• 📹 Session Recording: Capture terminal output for contextual help

🚀 Quick Start

Install

curl -LsSf https://raw.githubusercontent.com/kaleab-ayenew/vity/main/install.sh | sh

Configure

vity config

You'll be prompted for:

• Base URL: Your LLM provider endpoint
• API Key: Your API key (use NONE for local models)
• Model: Model name to use
• History Limit: Lines of terminal history to send (default: 1000)

Use

# Generate commands
vity do "find all python files larger than 1MB"
vity do "kill process using port 3000"

# Chat with AI
vity chat "explain this error message"
vity chat "what does chmod 755 do?"

# Use with context
vity record    # Start recording session
# ... work normally ...
vity do "fix this error"  # AI sees your terminal history
exit          # Stop recording

🔧 Provider Configuration Examples

OpenAI

Base URL: https://api.openai.com/v1
API Key: sk-your-openai-key
Model: gpt-4o-mini

Google Gemini

Base URL: https://generativelanguage.googleapis.com/v1beta
API Key: your-gemini-key
Model: gemini-1.5-flash

Ollama (Local)

Base URL: http://localhost:11434/v1
API Key: NONE
Model: llama3.2:3b

Other Providers

Works with any OpenAI-compatible API (Anthropic, Together AI, etc.)

📋 Requirements

• Python: 3.9+
• OS: Linux or macOS
• LLM Provider: OpenAI, Gemini, Ollama, or compatible API

🎯 Commands

Command	Description
vity do "<task>"	Generate shell command
vity chat "<question>"	Chat with AI
vity record	Start recording session
vity status	Show recording status
vity config	Manage configuration
vity config --reset	Reset configuration
vity install	Install shell integration
vity reinstall	Reinstall shell integration
vity uninstall	Completely remove vity

🔄 Context Recording

For the best experience, use recording to give Vity context:

vity record          # Start recording
# ... work normally, encounter errors ...
vity do "fix this"   # AI sees your terminal history and errors
exit                 # Stop recording

🗑️ Uninstalling

# Remove everything (shell integration, config, logs, chat history)
vity uninstall

# Or force without confirmation
vity uninstall --force

Then remove the package:

pipx uninstall vity  # if installed with pipx
# or
pip uninstall vity   # if installed with pip

🛠️ Troubleshooting

Command not found: Restart terminal or run source ~/.bashrc

API errors: Check your configuration with vity config --show

Reset everything: vity config --reset

🔒 Privacy

• Configuration stored locally in ~/.config/vity/
• Terminal history only sent during recording or with -f flag
• No data stored on external servers (except API calls)

📄 License

MIT License - see LICENSE file.

---

Need help? Run vity help or open an issue on GitHub.