Open Claw 101: Build-An-Agent Workshop
This event is FREE for members. Apply for a membership here :)
THE AGENDA
🕕 6:00PM — Doors Open
Arrive, grab a drink, and meet other builders.
🧠 6:30PM — OpenClaw 101: How Agents Actually Work
Note: you'll be given access to a virtual machine with OpenClaw already installed. Please bring your laptop.
A practical breakdown of how OpenClaw systems are structured:
how an OpenClaw agent differs from a one-shot chatbot
tools, workspace, and agent instructions
how memory works in OpenClaw by default
when to use one agent vs multiple agents
what’s real today vs what people overclaim about “agentic” systems
🎯 Outcome: You understand the OpenClaw model you’re about to build — and what makes it different from a generic LLM wrapper.
⚙️ 7:00PM — Guided Build: Your First OpenClaw Agent
You’ll enter a pre-configured OpenClaw environment and:
connect a model provider (such as OpenAI or Anthropic)
configure your OpenClaw workspace
define your agent’s role, instructions, and operating style
run the agent on a real task inside OpenClaw
🎯 Outcome: You leave with a working OpenClaw agent completing its first task.
🍕 7:45PM – 8:15PM — Food, Drinks & Compare Builds
Step away, recharge, and compare notes with other builders:
what are people building in OpenClaw?
what configurations are working well?
where are people getting stuck?
🎯 Outcome: New ideas, faster iteration, and a clearer sense of how others are approaching OpenClaw builds.
🔁 8:15PM — Turn It Into an OpenClaw System
Now we move from a single agent to a more capable OpenClaw setup:
improve the agent’s instructions and task boundaries
make better use of OpenClaw’s built-in memory and workspace
introduce additional agents where it actually helps
test simple delegation and multi-agent workflows
Note: in OpenClaw, memory is already part of the system by default, so this stage is about using it properly — not bolting it on afterwards.
🎯 Outcome: You see how an OpenClaw build evolves from a single agent into a usable agent system.
📢 9:00PM — Show & Tell
We’ll review a few builds live:
what worked
what broke
what changed once the system became more OpenClaw-native
how to improve the next iteration
🎯 Outcome: Learn faster by seeing how other people translated ideas into actual OpenClaw agent systems.