avavCommunityยท20 Agent Skills Included

harbor

Run and manage a complete local LLM stack with one command

Starts and manages 250+ containerized AI services like Ollama, llama.cpp, vLLM, Open WebUI, and ComfyUI with a single command. Eliminates manual Docker Compose wiring, model downloads, GPU configuration, and cross-service connection setup. Includes agent skills for testing, bug hunting, release automation, and building custom LLM proxy modules.
npx skills add av/harbor --all -g -y
Available:

Gives the AI agent a full CLI reference, safety rules (like avoiding hanging log commands), and workflows for configuring services, adding new ones, and running fact-driven development in this repository.

All Skills in This Repository (20)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install Harbor?โ–ผ

Run `npx skills add av/harbor --all -g -y` in your terminal to install all skills in this suite globally.

How to run LLMs locally with Harbor?โ–ผ

Run `harbor up` to start Open WebUI and llama.cpp pre-connected, then `harbor pull <model>` to download a model and `harbor open` to chat in your browser.

What services can Harbor manage?โ–ผ

Harbor manages over 250 AI services including backends like Ollama, vLLM, and llama.cpp, frontends like Open WebUI, and tools like SearXNG, ComfyUI, and Speaches, all pre-wired to work together.

Does Harbor work with AI coding agents?โ–ผ

Yes. `harbor launch` connects coding tools like Codex and Aider to your local models, and the included skills teach agents how to operate the Harbor CLI safely.

Do I need Docker experience to use Harbor?โ–ผ

No. Harbor handles all Docker Compose orchestration, configuration, and service networking for you, and `harbor doctor` checks your setup automatically.

Related Repositories in Software Engineering

View All in Software Engineeringโ†’