Skip to main content

Installation

The app ships as a self-contained desktop bundle (or a mobile build). Each desktop bundle carries its own backend (agent_core) and a bundled Python runtime, so there's nothing else to install to get started — no system Python, no separate server.

Download links

Release binaries are published on GitHub Releases. The exact URLs land on the Download page; the per-OS steps below assume you've downloaded the build for your platform.

macOS​

  1. Download the .app (Apple Silicon by default; a universal build that also runs on Intel Macs is available).
  2. Move it to Applications and launch it.
  3. On first launch macOS Gatekeeper may warn about an unsigned/ad-hoc build — right-click the app → Open to allow it.

The .app is fully self-contained: it embeds agent_core and a relocatable CPython, seeds your workspace on first run, and spawns its own backend.

Windows​

  1. Download and unzip the Windows bundle.
  2. Run the app executable in the unzipped folder (keep the folder intact — the backend and Python runtime live next to it).

Linux​

  1. Download and unpack the Linux bundle.

  2. Run the app binary inside bundle/:

    ./bundle/<app>

The bundle contains the app, agent_core, and a bundled python/ runtime. WebView support uses WPE WebKit — install your distro's WPE WebKit runtime if the app reports a missing webview.

Mobile (Android / iOS)​

Install the mobile build for your platform. On mobile the backend runs in-process (no separate server), so the experience is the same with nothing extra to install.

First run​

On first launch the app:

  1. seeds your workspace (notes, templates, the default agent),
  2. starts/uses the backend on http://127.0.0.1:9090,
  3. opens the editor.

Everything is local — your notes are markdown files in the workspace and structured data is local SQLite.

Enabling local AI models​

The AI features run on local models by default, so they work offline. The quickest path is Ollama:

# install Ollama (see ollama.com), then pull a model:
ollama pull qwen3.6:27b      # or any model you prefer

The app talks to Ollama on its default port. A bundled provider is also available, and if you'd rather use a hosted model you can point the app at any OpenAI-compatible endpoint in settings. Pick the provider and model in the app's model/settings screen.

Next​