How to Deploy SkillHub with Docker: Step-by-Step Guide

SkillHub gives enterprises a self-hosted registry for AI agent skills — but what does it take to actually get it running? The good news: the project ships with a docker-compose.yml plus dedicated server and scanner Dockerfiles, so you can be up in minutes. Here is the practical walkthrough.

Prerequisites

Step-by-Step Deployment

1. Clone the repository: git clone https://github.com/iflytek/skillhub and enter the directory.
2. Review the compose file: the included docker-compose.yml wires the server and scanner services. The server listens on port 8080.
3. Start the stack: run docker compose up -d. Both the API server and the skill scanner build from the provided Dockerfiles.
4. Create your first skill package: use the admin UI or CLI to publish a skill, assign a semantic version, and attach metadata such as description, tags, and required agent runtime.
5. Govern access: configure RBAC so only approved publishers can release new versions, and enable audit logging for compliance tracking.
6. Connect your agents: point Claude, Codex, Cursor, or any MCP-compatible agent at the registry endpoint and start discovering approved skills.

Pro Tips

SkillHub turns your agent skills into a governed, versioned, discoverable library. With the official Docker setup, deployment takes minutes — and the payoff in team productivity and compliance is immediate.