ShellGPT

ShellGPT (github.com/TheR1D/shell_gpt, MIT licensed, 12,000+ stars) is a Python command-line productivity tool powered by large language models. Installed via pip install shell-gpt, it generates shell commands from natural language, writes code snippets, produces documentation, and offers chat mode, an interactive REPL, function calling, reusable roles, and a request cache — across Linux, macOS, Windows, and all major shells (Bash, Zsh, PowerShell, CMD). By default it uses OpenAI's API (key required, usage-billed); it can also point at local models via Ollama.

The honest picture: it is one of the most-starred terminal AI tools for good reason — simple, scriptable, and genuinely useful for 'what was that ffmpeg incantation' moments. But OpenAI API costs apply per call, and the project's own docs warn it is not optimized for local models, so the free Ollama path may underwhelm. Newer agentic CLIs (Gemini CLI, Qwen Code) do more, but ShellGPT remains the lightweight single-purpose choice.

Run pip install shell-gpt. On first use you will be prompted for an OpenAI API key (stored in ~/.config/shell_gpt/.sgptrc). For local models, follow the Ollama setup guide in the project wiki.