Loading course…
Loading course…
Created by Shaunak Ghosh
Set up OpenClaw as a local-first, always-on personal assistant, control it from a real chat app, and deploy three daily workflows with a safety baseline. You will graduate into advanced operation, including multimodal usage, skill vetting and extension, and multi-agent routing with Docker sandbox isolation and evidence-based verification.
8 modules • Each builds on the previous one
Define OpenClaw’s core operating model: an agent-in-chat that uses tools/actions, with a strict verify-first loop (plan → act → evidence). Establish what “local-first always-on assistant” means in practice and where mistakes typically happen.
Complete installation/onboarding wizard, account + device pairing, and reach a first successful run that can answer a question and produce an evidence trail (logs/receipts).
Connect exactly one channel (WhatsApp/Telegram/Slack/Discord): pairing, allowlist, notification tuning, and basic commands. Emphasize safe defaults and confirmation boundaries in chat.
Learn the Control UI essentials needed for self-sufficiency: sessions, configuration profiles, logs, and the troubleshooting view. Focus on reading cause→effect from traces and undoing risky config.
Implement the “OpenClaw Daily Starter Pack” workflows with templates: (1) inbox triage with strict “draft vs send,” (2) calendar scheduling + meeting prep, (3) personal admin checklist (tasks/reminders/docs).
Discover, install, pin, and update skills. Learn a practical trust model: permissions requested, data access scope, maintainership signals, update hygiene, and rollback strategy.
Implement a concrete safety baseline: least privilege, read-only defaults, confirmations for write actions, secrets hygiene, and prompt-injection awareness. Turn these into enforceable settings and habits.
Graduate to advanced capabilities safely: multi-agent routing (work/personal/projects), always-on daemon mode, scheduled jobs/heartbeats, and media/voice/canvas toggles. Extend once with a simple MCP/custom skill workflow using guardrails + audit log, then lock reliability with “golden tasks” regression checks and evidence/receipt verification.
Begin your learning journey
In-video quizzes and scaffolded content to maximize retention.