Open-source AI app builder for creating and running agentic workflows with a visual drag-and-drop editor and multiple LLM providers.
brekkylab/ailoy
A library for building AI agents in Python, Node.js or Rust, where each agent gets its own microVM sandbox with GPU access.
- Stars
- 155
- Forks
- 7
- Open issues
- 26
- Last commit
- 1 hour ago
What is it?
Ailoy is a library for building AI agents inside your own code. Its central idea is giving each agent its own virtual machine, so the agent can install software, create and run scripts, and work in a general-purpose computing environment instead of being limited to predefined tool calls. The host system is only touched through folders and resources you explicitly mount or expose.
The VMs are based on krun-based virtualization and need no separate VM daemon. The virtualization supports the GPU, so models or even games can run inside an agent's VM. Models are reached either through provider CLIs such as Claude Code, Codex or Gemini CLI, or through provider APIs with a key; Amazon Bedrock, OpenRouter and vLLM are also listed as backends.
The library has Python, JavaScript/TypeScript and Rust APIs and runs on Linux, Windows and macOS. The repository includes examples such as CAD modeling, SQL analysis of a leaks database, and an agent that plays OpenTTD. The project is under active development and its API may change between versions.
Key features
- 01Gives each agent its own microVM sandbox to install software and run code.
- 02Supports GPU use inside the agent's VM, including Vulkan and Metal.
- 03Offers Python, JavaScript/TypeScript and Rust APIs.
- 04Connects to models through provider CLIs or API keys.
- 05Runs on Linux, Windows and macOS.
- 06Supports both complete-message and token-streaming output.
Use cases
- Building agents that write and render CAD models and check the results visually.
- Letting an agent analyse a dataset with SQL and Python it writes itself.
- Running ML models such as segmentation or text-to-speech on a guest GPU.
- Prototyping agents that operate a desktop application or game.
Quick start
- 1
Install and log in to a model CLI (Claude Code, Codex or Gemini CLI), or set a provider API key
- 2
Install the Python package
pip install ailoy-py - 3
Build and run an agent with the Python example code from the README
Things to know
- Ailoy is under active development and its API may change between versions.
- Running agents with a GPU is needed for the ML-based examples.
- Mounting host folders on macOS or Windows requires FUSE-T or Dokany.
- On Windows, the micro-VM needs the Windows Hypervisor Platform feature enabled.
In collections
Topics
FAQ
- What is ailoy?
- A library for building AI agents in Python, Node.js or Rust, where each agent gets its own microVM sandbox with GPU access. Ailoy is a library for building AI agents inside your own code. Its central idea is giving each agent its own virtual machine, so the agent can install software, create and run scripts, and work in a general-purpose computing environment instead of being limited to predefined tool calls. The host system is only touched through folders and resources you explicitly mount or expose.
- Is ailoy free to use? What license does it have?
- ailoy is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
- Is ailoy still maintained?
- Actively maintained. The last commit to ailoy was on Oct 22, 2026.
- What can I use ailoy for?
- Building agents that write and render CAD models and check the results visually. Letting an agent analyse a dataset with SQL and Python it writes itself. Running ML models such as segmentation or text-to-speech on a guest GPU. Prototyping agents that operate a desktop application or game.
- How do I install ailoy?
- Install and log in to a model CLI (Claude Code, Codex or Gemini CLI), or set a provider API key Install the Python package: pip install ailoy-py Build and run an agent with the Python example code from the README
- Where can I get ailoy?
- The source code and installation instructions are on GitHub: https://github.com/brekkylab/ailoy
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A minimal AI software engineering agent that solves GitHub issues or helps in the command line, using only bash and about 100 lines of core Python.
Embeddable, portable microVMs in Rust for running AI agents and untrusted code locally with hardware isolation.
A Python CLI that wraps coding agents like Codex and Claude Code with an optional supervisor, reviewer, adversary, and local log distiller, each on its own model.
Open-source infrastructure for running AI-generated code in isolated cloud sandboxes, with JavaScript and Python SDKs.
A Python agent runtime that wraps an LLM with tools, memory, a policy on every action, a sandbox for commands, crash-safe runs and a full trace of each step.