veadk-skills

What This Skill Does

The veadk-skills tool is a specialized OpenClaw agent capability designed for developers working within the VeADK ecosystem. It acts as an intelligent generator and converter, streamlining the process of creating, migrating, and optimizing AI Agents. Whether you are building from scratch or porting complex logic from existing frameworks like LangChain, LangGraph, or Dify, this skill automates the architectural mapping, prompt engineering, and boilerplate generation required for production-ready VeADK code.

Installation

To integrate this skill into your OpenClaw environment, execute the following command in your terminal:

clawhub install openclaw/skills/skills/yaozheng-fang/veadk-skills

Ensure your local development environment is configured to allow file-write permissions, as the skill automatically generates project structures, including __init__.py and agent.py files using the save_file.py script provided by the framework.

Use Cases

1. Greenfield Agent Development: Rapidly scaffold new VeADK agents by providing a natural language functional requirement. The skill performs deep analysis to suggest the optimal system architecture and refined system prompts.
2. Legacy Migration: Convert existing LangChain or LangGraph chains into optimized VeADK-compliant code, ensuring that dependencies and logic flows adhere to the VeADK architecture standards.
3. Dify Transformation: Import complex Dify workflow DSL files and transform them into programmable, flexible VeADK agent implementations, preserving original workflow logic while enhancing scalability.

Example Prompts

1. "I need to build a customer support agent using VeADK that handles ticket classification and routing. Please analyze the requirements and generate the base agent structure."
2. "I have an existing LangGraph workflow designed for document extraction. Please convert this logic into a VeADK agent implementation, mapping the state nodes to the VeADK coding standards."
3. "Convert this attached Dify YAML workflow into a fully functional VeADK agent agent.py file."

Tips & Limitations

• Pre-requisites: Ensure your documentation references in references/generator/ and references/converter/ are up to date, as the agent relies on these internal docs to ensure code quality.
• File Management: The skill triggers automated file writing; verify the directory permissions before execution to prevent access errors.
• Complexity: While the skill is robust for standard workflows, highly custom or proprietary third-party integrations within LangChain might require manual review post-conversion to ensure all Python dependencies are correctly resolved.