other / agent

Rakazo

A self-hosted AI agent app with persistent bots, computers for browser, terminal and files, and clients for web, desktop, iPhone and Android.

At a glance

Purpose
A self-hosted AI agent app with persistent bots, computers for browser, terminal and files, and clients for web, desktop, iPhone and Android.
Runs in
Local agent with shell, Web service
Operating systems
Not stated
Needs
Runtime: Docker Engine, Compose, curl and OpenSSL; Needs internet: Not stated; Needs accounts or API keys beyond your agent: Not stated
Files and data
Reads files: Yes; Writes files: Yes; Sends data to other services: Not stated
Tool type
Agent
Delivery
From GitHub
Version
Latest on the main branch
Access
Code on GitHub
Compatibility
According to the README: web, an Electron desktop app, and iPhone and Android clients; Docker Engine, Docker Desktop and OrbStack for local computers; remote computer providers E2B, Daytona, CreateOS and Box; Composio, Pipedream Connect, MCP, OpenAPI and Treg for integrations; voice keys from ElevenLabs, OpenAI, Cartesia and Fish Audio; and images for linux/amd64 and linux/arm64. The README gives Caddy as its HTTPS setup for a single VM.
External costs
According to the README, you bring your own model credentials and computer provider, or run the whole stack locally. The remote computer providers E2B, Daytona, CreateOS and Box each need their own API key. Voice mode uses a key from ElevenLabs, OpenAI, Cartesia or Fish Audio. The Composio and Pipedream Connect catalogs are optional and use their own credentials. The README says Treg is usage-metered and that self-hosters supply their own Treg token. It gives no prices. The providers you connect may charge for usage.
Permissions
The README says bots have browser, terminal, file and graphical desktop access through shared Team Computers or isolated Private computers, with local Docker computers enabled by default and remote providers requiring their own API keys. The README describes model credentials, voice-provider keys and integration credentials. The README gives a downloaded installer script executed with bash that downloads Compose files, writes random secrets to .env and starts services. The README says the desktop This computer route installs published images under the app's data directory and leaves the stack running after the app quits; Stop Local Stack stops it. The README says Docker sandboxes need a supervisor token. The README warns that docker compose down -v deletes all PostgreSQL state and says to recreate the volume only after a backup or when the data is disposable. The README does not describe approval prompts or permission modes for bot actions.
Usage terms
Apache License 2.0, as stated in the repository. See its LICENSE file for the exact terms. The README says Treg integration terms require a written agreement for hosted resale.
Interface
web

Code on GitHub

elie222/rakazo

Apache-2.0

TypeScript3,810 stars657 forksLast pushed 12 Oct 2026 UTC

Open-source Grok Bot alternative. Choose your own model and sandbox.

Topics: ai-agents chatgpt docker electron expo grok llm self-hosted

Public GitHub data fetched · 0 days old

Skill folders in this repository (2)

Folders that contain a SKILL.md file, from GitHub's file list on .

View on GitHub ↗

What it does

The README describes Rakazo as an open-source, self-hosted AI assistant for persistent teammates. The README lists web, Electron desktop, iPhone and Android clients, labels it beta, and says users bring model and computer credentials or run the stack locally.

The README lists persistent bots with conversations, memory, routines and history, plus voice replies, dictation and calls. The README describes shared Team Computers and isolated Private computers with browser, terminal, file and graphical desktop access, and says bots can delegate to peers or short-lived subagents.

The README says local Docker computers are enabled by default and names E2B, Daytona, CreateOS and Box as remote providers. The README lists Composio, Pipedream Connect, MCP, OpenAPI and Treg integrations, and describes clients connecting to the same API.

Who it is for

People who want a self-hosted assistant with persistent bots and agent computers, and developers who want to work on its code.

Example tasks

Install Rakazo from the published images with the installer script, open it in a browser and create an account.
Connect a model during onboarding and create your first bot.
Run the backend on a server and connect the desktop app to it with Existing instance.
Add an HTTPS MCP server, a Treg endpoint or an OpenAPI JSON document from Integrations.
Use voice mode to dictate to a bot or call it, with your own voice provider key.

Getting started

The README's quick start needs Docker Engine, the Compose plugin, curl and OpenSSL, with no clone or Node install. You make a folder, download the installer script with curl and run bash install-images.sh. The README says the installer downloads the Compose files, creates .env with random secrets and starts Rakazo. Then open the local address in a browser, create an account and connect a model. For a server, run bash install-images.sh --prepare-only, edit .env for the sandbox provider, its API key and your host name, and run bash install-images.sh again. The README also describes a source checkout route, the desktop app, and a setup prompt file for an agent-assisted install. The README has the exact commands.

Limitations

According to the README, the published-image quick start needs Docker Engine, the Compose plugin, curl and OpenSSL. Working from a source checkout needs Node.js 22.22.2 or newer in the 22.x line, Node.js 24.x or Node.js 26 and later, pnpm 9 and Docker. The README says Node.js 23.x and 25.x are not supported. You connect a model during onboarding or set a key such as OPENROUTER_API_KEY. The README says voice mode uses a key you bring from ElevenLabs, OpenAI, Cartesia or Fish Audio, and remote computers need the matching provider API key. For a public server, the README says to put HTTPS in front of port 5173. The desktop app needs Docker Desktop, OrbStack or Docker Engine to run Rakazo on the same computer. The README says Rakazo is in beta. The repository can change at any time; check its README before you install.

Data practices

The README says users can run the stack locally or configure external model and computer providers. The README describes PostgreSQL storage and says the default Compose database is network-internal. The README says connector credentials are encrypted on the server. The README describes Composio, Pipedream Connect and user-added remote MCP, Treg and OpenAPI sources; data sent to external services depends on the configured providers and integrations. The README does not describe telemetry or analytics.

Community experience

Not rated yet

Sign in to share your experience.

Comments 0

No comments yet.

Sign in to comment

Report a listing concern