One hands-on course, three days: drive one agent well, teach it your standards, then multiply it into a production system. Every learner leaves with a working personal toolkit — CLAUDE.md, a custom skill, a hook, a custom command, and a published plugin — plus a capstone multi-agent build with a green CI check. Fully refreshed July 2026.
Twelve labs chain into one final artifact — your personal Claude Code toolkit repo.
Do this before Day 1 — license check, native installer, /doctor green
Seat/license check, install per OS, sample repo, and the "you are ready" marker
⬇ Download PDFThe agentic loop, memory, context, checkpoints & rewind, Auto mode, cost & permissions
Native install, Sonnet 5 model lineup, the loop, memory, context, permissions, git
⬇ Download PPTXBuild a dashboard from a single prompt — plus checkpoints & rewind
⬇ Download PDFCLAUDE.md policy files, /init, and auto memory in action
⬇ Download PDFThree recovery drills built on the checkpoint/rewind system
⬇ Download PDFPermission modes incl. Auto, deny→ask→allow rules, git workflow to PR
⬇ Download PDFSkills, hooks, custom commands, plugins & marketplaces
SKILL.md, progressive disclosure, hooks, commands, plugin trust model, marketplaces
⬇ Download PPTXA SKILL.md that triggers reliably — current frontmatter fields
⬇ Download PDFPick from 27 hook events, wire a PostToolUse hook + a $ARGUMENTS command
⬇ Download PDFSkill evals, A/B benchmarking, and debugging skills that won't load
⬇ Download PDFclaude plugin init — package your toolkit as a plugin and publish to a team marketplace
⬇ Download PDFMCP, nested subagents & agent teams, headless CI/CD, scale & cost
MCP, nested subagents, agent teams, claude -p in CI/CD, caching & worktrees
⬇ Download PPTXCustom subagents, nested delegation, a grader review pass, and agent teams
⬇ Download PDFMulti-agent build on your toolkit repo + a green headless CI check
⬇ Download PDFPrint these out and keep them handy
Reference implementations and examples
Configuration files for different project types — Python, TypeScript, enterprise, security-focused
↗ View on GitHubReady-to-use slash commands — commit, PR review, refactor, docs, security scan
↗ View on GitHubMaterials for the Claude.ai business-user course live on their own page.
Claude AI for Business →