av/harbor
Stop configuring your AI stack. Start using it. One command brings a complete pre-wired LLM stack with hundreds of services to explore.
What it solves
Harbor eliminates the complex manual setup required to run local AI stacks. Instead of individually installing and configuring backends, frontends, and supporting tools, it automates the entire process so you can start using local LLMs, web search, voice chat, or image generation immediately.
How it works
Harbor acts as a command-line interface (CLI) and companion app that manages Docker Compose orchestration for you. When you run a command like harbor up, it automatically handles the configuration and network connectivity between services. It pre-wires components—such as connecting a chosen LLM backend (like Ollama or vLLM) to a frontend (like Open WebUI)—so they work together out of the box. You can also use it to inject local models into existing coding tools or script complex agentic workflows.
Who it’s for
It is designed for developers and AI enthusiasts who want to run local AI infrastructure without spending time on system configuration, networking, or dependency management. It is particularly useful for those who want to experiment with different inference engines, agentic coding tools, or RAG setups on their own hardware.
Highlights
- Single-command deployment for complete AI stacks using
harbor up. - Supports a wide range of backends, frontends, and satellite services like SearXNG, ComfyUI, and Speaches.
- Built-in support for agentic coding workflows via Harbor Boost, including web research and deliverable audits.
- Easy integration with host-native coding tools like VS Code, Claude, and Copilot.
- Flexible "eject" feature that exports your current configuration to a standalone Docker Compose file if you want to move away from Harbor.
- Includes tools for remote access, such as QR code generation for mobile and built-in tunneling for internet exposure.
Related
- Dispatch
- Project
- Project
- Project
- Project