About this site
- What it is
- AgentHub is a learning site focused on OpenClaw and AI agent workflows—not the official OpenClaw product site.
- What you'll find
- Tutorials, workflow patterns, and concrete examples. Technical explanations for developers.
Learn OpenClaw and AI agent workflows
Technical tutorials, concepts, and workflow examples for developers building multi-step AI systems with OpenClaw.
OpenClaw is a framework for building AI agent workflows. It lets you define agents that reason and call tools, then chain them into workflows—sequential, parallel, or branching. Use it to automate multi-step tasks that need search, synthesis, or decision-making.
Learn more in What is OpenClaw. Understand the architecture in How OpenClaw works. See patterns in OpenClaw workflow examples.
Where to start
These pages cover the essentials. Begin with an overview, then dive into concepts and examples.
What is OpenClaw?
An overview of OpenClaw: what it is, its core concepts, and how it fits into the AI agent ecosystem.
Read more →How OpenClaw works
Learn the architecture: agents, tools, workflows, and how they connect. Understand the mental model before writing code.
Read more →OpenClaw workflow examples
Concrete examples: sequential chains, parallel execution, branching, and human-in-the-loop patterns with code.
Read more →
OpenClaw tutorials and guides
Step-by-step tutorials for common workflow use cases. Each guide explains the design and walks through implementation.
Research assistant workflow
A multi-agent setup that searches, summarizes, and synthesizes information from multiple sources.
Read more →Customer support routing
Route tickets to specialized agents by intent. Handle escalation to humans when needed.
Read more →Content creation pipeline
Draft, edit, and review content with separate agents. Parallelize steps where possible.
Read more →
Topic hub
Navigate to learning pages by category.