OpenClaw Foundation teaches attendees how to deploy, configure, and operate OpenClaw across two hands-on days. OpenClaw is the open source, self-hosted gateway for AI agents across chat platforms and web interfaces. It enables teams and individuals to connect messaging channels, such as Slack, Telegram, Discord, and WhatsApp, to agent-based workflows running through a single gateway process. In this course, participants learn how OpenClaw handles sessions, routing, workspaces, skills, and model connectivity. They also see how it manages multi-agent isolation while exposing AI assistants through practical communication surfaces.
Through lecture, demonstrations, and extensive guided labs, attendees install OpenClaw and complete onboarding. They configure providers, connect channels, and work with the browser dashboard and control surfaces. The course then advances into skills, workspace structure, agent bindings, multi-agent routing, and operational controls for safe deployment. By the end of the class, participants can build and manage OpenClaw environments that are useful, secure, and ready for team or personal use.
Specific labs walk through configuring ~/.openclaw/openclaw.json, validating the Control UI, and connecting a channel end-to-end so messages flow from a real chat platform into an agent session. Later labs add a workspace skill and observe how the change alters agent behavior, then extend a single deployment into several isolated agents bound to different channels, accounts, and team contexts.
A dedicated security module covers sandboxing, tool allowlists, mention-based approval models, and safe defaults for personal, team, and public-facing agents. A troubleshooting lab has participants diagnose and repair a broken configuration, reinforcing how routing, sessions, and channel auth fit together. The class closes with a capstone lab where attendees assemble a small production-style deployment, weighing local, server-hosted, and hybrid tradeoffs along the way.
Who Should Attend
Developers, Platform Engineers, DevOps personnel, SREs, Architects, AI Engineers, Technical Managers
What Attendees Will Learn
Upon completing OpenClaw Foundation, attendees will be able to:
- Install and configure OpenClaw for local and remote operation
- Connect OpenClaw to model providers and supported messaging channels
- Build and manage isolated agent workspaces, skills, and routing configurations
- Design multi-agent OpenClaw deployments with channel bindings and per-agent separation
- Apply security, sandboxing, approval, and governance controls to OpenClaw environments
- Troubleshoot sessions, routing, and channel integrations in real-world deployments
Prerequisites
Participants should have basic command-line skills, familiarity with JSON configuration files, and a general understanding of modern LLM-based assistants.