ASB is an intelligent message bus that evolves traditional ESB patterns to enable advanced AI Agent service orchestration and discovery.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install asb
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install asb using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
ASB (Agent Service Bus) is a sophisticated platform designed to bridge the gap between traditional Enterprise Service Bus (ESB) architectures and modern AI Agent ecosystems. By leveraging the Model Context Protocol (MCP), ASB allows developers to integrate powerful orchestration, governance, and service discovery capabilities directly into environments like Claude Desktop. It provides a centralized hub where multiple agents and services can interact, making it a cornerstone for teams building complex multi-agent workflows.
As a primary entry point for Openclaw Skills, ASB facilitates Agent-to-Agent (A2A) collaboration. It treats AI capabilities as discoverable services, allowing for dynamic task routing, smart load balancing, and comprehensive monitoring of agentic processes. Whether you are running locally via CLI or integrating through an IDE, ASB ensures that your agent infrastructure remains scalable and manageable.
To get started with ASB, ensure you have uv and python installed on your system. Follow these steps to clone and sync the environment:
# Clone the repository
git clone https://github.com/qiyueqiu/Agent-Service-Bus.git
cd Agent-Service-Bus
# Install dependencies using uv
uv sync
To integrate with Claude Desktop, add the following to your claude_desktop_config.json:
{
"mcpServers": {
"asb": {
"command": "uv",
"args": [
"--directory",
"/path/to/Agent-Service-Bus",
"run",
"asb",
"mcp",
"serve",
"--transport",
"stdio"
]
}
}
}
ASB organizes its ecosystem through a structured metadata taxonomy and tool-based schema. Below are the core functional categories:
| Component | Key Tools | Data Managed |
|---|---|---|
| Core ASB | asb-service-discovery, asb-service-composition | Service registries, orchestration manifests, and governance logs. |
| Utility | calculator, echo, message-send | Transient execution data and communication payloads. |
| Provider | llm-provider | API configurations and model response metadata. |
| A2A Integration | AgentCard | JSON-based capability definitions and agent identification schemas. |
Loading
An AI-powered skill that transcribes video audio and burns styled subtitles directly into the video file for social media and accessibility.

An advanced web extraction tool that renders JavaScript-heavy websites and returns clean Markdown optimized for LLM processing.

An AI-powered vision skill that identifies and highlights specific delivery packages on crowded shelves using pickup codes.

An intelligent fitness logging skill that parses natural language workout data into a local MySQL database for progress tracking and smart training plans.

An AI-driven video assistant designed for cemeteries and memorial parks to create pre-planning educational content and digital grounds tours.

A massive collection of 34 production-tested Openclaw Skills covering every major category from finance and productivity to health and education.








































