What is the purpose of the visual diagram?

It provides a flow-based representation of the agent's logic using HTML and CSS, making it easier to verify the architecture before coding begins.

How do I start building a new agent?

Use the /agent-builder command to initiate an interactive interview where Claude will gather your requirements and guide you through the process.

Can I customize the LLM used by the generated agent?

Yes, you can configure the model using the LLM_BASE_MODEL environment variable, though it defaults to Claude 3.5 Sonnet.

Does this skill provide the actual code?

Yes, after generating and approving a technical plan, the skill generates a complete Python implementation in your output directory.

Where does the skill get its SDK information?

The skill dynamically clones and analyzes the latest documentation and examples directly from the official OpenHands repositories to ensure accuracy.

Agent SDK Builder

Name: Agent SDK Builder
Author: OpenHands

byOpenHands

•

Data Science & ML

Builds custom AI agents using the OpenHands SDK through an interactive interview and guided implementation process.

The Agent SDK Builder skill simplifies the creation of specialized AI agents by guiding users through a structured requirement-gathering interview. It automatically researches the OpenHands Software Agent SDK documentation, analyzes code examples, and generates both a comprehensive technical plan and a visual workflow diagram. Once the design is approved, it produces production-ready Python code, allowing developers to rapidly prototype and deploy sophisticated agents with built-in logging and configurable LLM settings.

Key Features

01Interactive interview process to define agent requirements and scope

0240 GitHub stars

03Interactive visual flow diagrams for architectural clarity

04Automated Python code generation for the finalized agent implementation

05Generation of detailed technical implementation plans (SDK_PLAN.md)

06Automatic research of OpenHands SDK documentation and code examples

Use Cases

01Translating high-level agent ideas into technical SDK implementations

02Visualizing complex agent workflows before writing production code

03Rapidly prototyping a new AI agent for specific software engineering tasks

Key Features

01Interactive interview process to define agent requirements and scope

0240 GitHub stars

03Interactive visual flow diagrams for architectural clarity

04Automated Python code generation for the finalized agent implementation

05Generation of detailed technical implementation plans (SDK_PLAN.md)

06Automatic research of OpenHands SDK documentation and code examples

Use Cases

01Translating high-level agent ideas into technical SDK implementations

02Visualizing complex agent workflows before writing production code

03Rapidly prototyping a new AI agent for specific software engineering tasks