LLM Cmd 2

Inspired by llm-cmd

LLM Cmd 2 generates shell commands based on natural language prompts. The output is loaded into the user's shell:

You can also use llm-cmd to fix commands that resulted in an error, without having to retype anything:

LLM Cmd 2 iterates on the original package:

• Includes OS information in prompts for more accurate command generation.
• Loads output into the shell rather than executing directly, so commands appear in history.

Installation

1. First, install the llm CLI tool
2. Run the installation script:

   curl -fsSL https://raw.githubusercontent.com/tombedor/llm-cmd2/refs/heads/main/install.sh | /bin/bash

3. Restart your terminal or run:

   source ~/.zshrc

Usage

llmcmd "your natural language request"

Examples

# Find files
llmcmd "find all Python files in current directory"

# Process management
llmcmd "kill process on port 3000"

# File operations
llmcmd "compress this folder to a zip file"
llmcmd "extract archive.tar.gz"

# Git operations
llmcmd "undo last git commit"
llmcmd "create new branch called feature-auth"

# System information
llmcmd "show disk usage"
llmcmd "list running processes sorted by memory usage"