Code with AI, entirely local
Today we are releasing Rune—a command-line coding assistant that brings SAGEA's powerful SAGE models to your local machine. Rune allows you to use natural language to explore, modify, and interact with your projects through a comprehensive toolset, all running privately on your own hardware.
Rune uses Ollama as the local runtime to execute SAGE models efficiently. Your code never leaves your computer.
Installation
Getting started with Rune takes a single command. The installer detects your platform and architecture, downloads the latest binary, and sets everything up automatically.
# One-line install
curl -fsSL https://sagea.space/rune-cli/install.sh | bash
What Rune can do
Rune is built for real development workflows. It can read, write, and patch files, execute shell commands, search your codebase recursively, manage task lists, ask you clarifying questions, and even delegate work to subagents—all from a single interactive chat prompt.
It automatically scans your project's file structure and Git status to provide relevant context, and ships with multiple built-in agent profiles tailored for different workflows: planning, code refactoring, or fully autonomous execution.
Key features
- 100% Local & Private — All processing happens on your machine. Your code never leaves your computer.
- Powered by SAGE — Leverages SAGEA's state-of-the-art reasoning models for deep understanding of code and logic.
- Interactive Chat — Conversational AI agent that reasons about your requests and breaks down complex tasks.
- Powerful Toolset — File manipulation, code searching, version control, and command execution from one prompt.
- Subagent Delegation — Delegate tasks to specialized subagents for parallel, context-efficient work.
- Skills System — Extend Rune with custom slash commands and reusable skill definitions.
- Highly Configurable — Customize models, tool permissions, and UI preferences via
config.toml.

