lumi-installer

<p align="center"> <img src="docs/lumi.png" alt="Lumi" width="160" /> </p> <h1 align="center">Lumi</h1> <p align="center"> She reads code. She writes it too. She remembers. She learns. </p> <p align="center"> <a href="#install"><img alt="Install" src="https://img.shields.io/badge/install-curl%20%7C%20bash-0aa?style=flat-square" /></a> <a href="https://github.com/ankurCES/lumi-installer/wiki"><img alt="Wiki" src="https://img.shields.io/badge/docs-wiki-success?style=flat-square" /></a> <img alt="Platform" src="https://img.shields.io/badge/platform-macOS%20%7C%20Linux%20%7C%20Windows-lightgrey?style=flat-square" /> <img alt="Status" src="https://img.shields.io/badge/status-alpha-orange?style=flat-square" /> <img alt="Made with Electron" src="https://img.shields.io/badge/made%20with-Electron-2c3e50?style=flat-square&logo=electron&logoColor=white" /> <img alt="Built for Claude" src="https://img.shields.io/badge/built%20for-Claude%20Code-d97757?style=flat-square&logo=anthropic&logoColor=white" /> </p> <p align="center"> <img src="docs/moonrunner.gif" alt="Moonrunner — Lumi's animated banner" width="720" /> </p> --- ## About **Lumi** is a desktop AI workspace — your AI coworker for code, documents, and image / video generation, with team mode for long-running multi-step work. - **Optimised for Claude.** Bring your Claude Code CLI as the runtime; Lumi auto-provisions a sandboxed overlay (skills, MCP servers, persona) that never touches your global `~/.claude/` config. - **Built-in Lumi Engine** runtime. Works with Anthropic, OpenAI, Google Gemini, and locally-hosted models via Ollama, LM Studio, and similar providers. Each agent picks its own model. - **ComfyUI bundled in.** Image and video generation pipelines are auto-provisioned on first launch — no separate install. - **Team mode** for long-running tasks. Spawn a lead agent + a configurable team of specialists; the lead delegates, tracks progress, and reports back. --- ## Install The installer handles everything: build-tool prerequisites (Bun, Node, pnpm, gh CLI), source clone, dependency install, app build, and first launch. Expect 3–5 minutes end-to-end. ### macOS / Linux ```bash curl -fsSL https://raw.githubusercontent.com/ankurCES/lumi-installer/main/install-from-source.sh | bash ``` ### Windows (PowerShell) ```powershell irm https://raw.githubusercontent.com/ankurCES/lumi-installer/main/install-from-source.ps1 | iex ``` You'll be asked whether to perform a clean install or an update, and prompted for GitHub authentication (gh CLI, PAT, or SSH — auto-detected). The Ink-based TUI guides you through the rest. --- > **Note — alpha release.** Lumi is currently in private alpha. The install will ask for a GitHub access token (or `gh` CLI auth) because the source repo is not public yet. If you'd like access, reach out and a time-limited PAT will be shared with you. The token, the build pipeline, and the installed app all run locally on your machine — nothing is sent to a remote service.