Setup & Installation
openclaw skills install @kenblive/communicate
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
npx clawhub install communicate
What This Skill Does
ModelReady starts a local or Hugging Face model as an OpenAI-compatible endpoint and lets you chat with it directly from a conversation. It handles server startup, status checks, and teardown without switching tools.
It removes the manual steps of configuring vLLM, managing endpoints, and opening a separate chat interface.
When to use it
- Testing a newly downloaded Hugging Face model without writing server code
- Running a quantized local LLM on a specific port during development
- Chatting with a self-hosted model from the terminal
- Switching between multiple model servers by port
- Exposing a local model to other tools via OpenAI-compatible API
Example Workflow
Here's how your AI assistant might use this skill in practice.
INPUT
User asks: start Qwen2.5-7B on port 8010 and send it a test message
AGENT
- 1Run /modelready start repo=Qwen/Qwen2.5-7B-Instruct port=8010
- 2Poll /modelready status port=8010 until the server reports ready
- 3Run /modelready chat port=8010 text="hello"
- 4Return the model response to the user
OUTPUT
The locally running Qwen2.5-7B instance replies to the greeting
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.
URL environment variable set to the target host address