Shell-Pilot
FreeInteract with LLM using Ollama models(or openAI, mistralAI)via pure shell scripts on your Linux(or MacOS) system, enhancing intelligent system management without any dependencies.
FreeFree tier
Inputs: textOutputs: text
About Shell-Pilot
Shell-Pilot is a lightweight, dependency-free shell script that enables interaction with multiple large language models—including OpenAI, Ollama, Mistral AI, LocalAI, ZhipuAI, Anthropic, Moonshot, and Novita AI—directly from the terminal. It supports local and remote models, session persistence, command generation, and a modular plugin system for enhanced system management. The script is pure shell (bash) and only requires curl and jq, making it easy to install on Linux and macOS.
Key Features
Supports multiple LLM providers: OpenAI, Ollama, Mistral AI, LocalAI, ZhipuAI, Anthropic, Moonshot, Novita AI
Pure shell script with no dependencies (only curl and jq required)
Session persistence and chat context for remembering conversation history
Pass input via pipe/redirector, script parameter, or normal chat mode
Generate commands and run them in terminal (code chat mode)
Modular plugin design for easy expansion and customization
View history and list all available models
Easy config management with vi mode (option e)
Enhanced system interaction: check package versions, manage aliases from command line
Pros & Cons
Pros
- Lightweight and dependency-free (pure shell script)
- Supports a wide range of LLM providers, including local options
- Session persistence for context-aware, multi-turn conversations
- Modular plugin architecture makes it extensible and adaptable
- Easy installation via curl and simple configuration
- Works on both Linux and macOS
Cons
- Not production-ready; intended for testing and personal use only
- Requires external dependencies: curl and jq must be installed
- Limited to terminal-based interaction (no GUI)
- Setting up local models (Ollama/LocalAI) requires additional steps
Best For
Intelligent system management from the terminal using natural languageQuick command generation and execution without leaving the shellLocal LLM integration with Ollama for private, offline useCode chat and development assistance for generating code snippetsBasic system administration tasks like checking package versions and managing aliases
FAQ
What are the prerequisites for using Shell-Pilot?
You need curl and jq installed on your system, and an API key for the desired LLM provider (e.g., OpenAI). Optionally, you can install glow for Markdown rendering.
How do I install Shell-Pilot?
Run the install script from the repository: curl -sS -o spilot_install.sh https://raw.githubusercontent.com/reid41/shell-pilot/main/spilot_install and follow the prompts to enter your API key.
Can I use local models with Shell-Pilot?
Yes, Shell-Pilot supports local models via Ollama and LocalAI. You need to have Ollama or LocalAI installed and running, then configure Shell-Pilot to use them.
Is Shell-Pilot suitable for production environments?
No, the project is a test project and is not production-ready. Do not use it in critical or production systems.