Instantly transform local or Hugging Face models into OpenAI-compatible endpoints for direct chat interaction within your agent environment.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install communicate
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install communicate using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is ModelReady?
ModelReady is a high-performance utility within the Openclaw Skills ecosystem designed to bridge the gap between model repositories and interactive agents. By leveraging vLLM, it allows developers to serve models from local storage or Hugging Face repositories as standardized APIs without leaving their terminal or chat interface.
This skill eliminates the friction of manual server configuration, providing a seamless way to test, interact with, and integrate various LLMs into your development lifecycle. Whether you are working with private local weights or exploring the latest releases on Hugging Face, ModelReady ensures your models are always a single command away from being chat-ready.
ModelReady Use Cases
- Quickly deploying a local model for immediate testing and validation.
- Interfacing with Hugging Face repositories directly from a chat interface without manual setup.
- Testing model responses and behavior using standardized OpenAI-compatible endpoints.
- Running private, locally hosted LLMs to maintain data sovereignty while utilizing Openclaw Skills.
How ModelReady Works
- The user specifies a local file path or a Hugging Face repository ID using the start command.
- ModelReady initializes a local server instance using vLLM on a user-defined port.
- The server exposes a fully OpenAI-compatible API endpoint locally.
- Users send chat prompts to the active server directly through the agent interface to receive real-time responses.
- The skill provides management commands to check server status, stop instances, or reconfigure default host parameters.
ModelReady Setup
To get started with ModelReady, ensure your environment has bash and curl available. You can configure the service using the following steps:
# Start a model from Hugging Face on a specific port
/modelready start repo=Qwen/Qwen2.5-7B-Instruct port=19001
# Alternatively, start a local model with specific hardware optimizations
/modelready start repo=/path/to/local/model port=8010 tp=4 dtype=bfloat16
ModelReady Data Schema & Taxonomy
ModelReady organizes its operational metadata and server configurations using a structured parameter set. This ensures consistency when interacting with the Openclaw Skills framework:
| Parameter | Description | Requirement |
|---|---|---|
repo |
Path to local model or Hugging Face repository ID | Required |
port |
The network port the vLLM server will listen on | Required |
tp |
Tensor Parallelism degree for multi-GPU setups | Optional |
dtype |
Data type precision (e.g., bfloat16, float16) | Optional |
ip |
Default host IP address for the model server | Optional |
ModelReady Advanced Features
- Support for Tensor Parallelism (tp) to enable high-performance inference across multiple GPUs.
- Custom data type (dtype) configuration to optimize memory usage and precision for specific hardware.
- Persistent configuration for default IP and port settings to streamline repetitive workflows.
- OpenAI-compatible API formatting, allowing any tool that supports the OpenAI spec to interact with your Openclaw Skills managed models.
SKILL.md
Loading
Related Openclaw Skills
Weekly Skills Update
An automated maintenance skill that updates all agent capabilities and generates comprehensive versioning reports.
Sales Oratory Master
An AI-powered negotiation assistant designed to handle B2B sales objections through strategic value-shifting and diagnostic communication.
Narrative Voice
A narrative-driven dialogue skill that transforms standard AI responses into warm, Gaiman-inspired storytelling experiences.
Logic Hunter
A high-precision logic verification tool for Openclaw Skills that uses the Golden Triangle framework to hunt for truth and score evidence credibility.
A-Share Short-Term Decision
A data-driven decision engine for CN A-share momentum trading with a 1-5 day horizon.
Subskill Generation Rule
A structural protocol for AI agents to organize code into modular subskills and centralized data directories.
Featured*
