Full Playbook
For deep understanding. 14 modules across three tiers, ~19 hours. Theory + practice. Covers security, team adoption, CI/CD, and the mental models behind effective AI-assisted development.
Claude Code in a Day
12 hands-on modules, ~6 hours. No theory — just your codebase, your tools, and results. From install to multi-agent workflows in a single day.
Go deeper with the full playbook
Section titled “Go deeper with the full playbook”Full playbook: three tiers
Section titled “Full playbook: three tiers”Tier 1: Foundations
For everyone — developers, DevOps, and product owners. 5 modules, ~7 hours. No prerequisites.
Tier 2: Mastery
For developers and DevOps engineers. 5 modules, ~7 hours. Requires Tier 1 + 1–2 weeks practice.
Tier 3: Operations
For DevOps engineers and tech leads owning production systems. 4 modules, ~5 hours. Requires Tiers 1 & 2.
Useful links
Section titled “Useful links” Claude Code Cheat Sheet Quick-reference commands and shortcuts for Claude Code.
Claude Code in Action Anthropic's official Claude Code training course.
Claude Code Documentation Official Claude Code documentation from Anthropic.
Mastering Claude Code in 30 min Video walkthrough covering Claude Code essentials.
How each module works
Section titled “How each module works”Every module follows the same pattern:
- Pre-work (15–20 min) — Self-paced theory reading that builds a mental model
- Workshop (45–60 min) — Self-directed hands-on session applying the theory
- Takeaway — A concrete artifact, habit, or workflow you use from that day forward
Audience guide
Section titled “Audience guide”| Role | Recommended path |
|---|---|
| Product Owner | Tier 1 · M06 · M08 |
| Developer | Tiers 1 and 2 |
| DevOps / Tech Lead / Senior Developer | All three tiers |
Curriculum at a glance
Section titled “Curriculum at a glance” M01 — How LLMs Actually Work Transformer architecture, autoregression, hallucinations, and context windows.
M02 — Prompt Engineering Zero-shot, few-shot, chain-of-thought, meta prompting, and model selection.
M03 — Specs Are Source Code Plan Mode, spec templates, and the spec-first paradigm shift.
M04 — Context Engineering The four failure modes of large contexts and how CLAUDE.md prevents them.
M05 — Agents and MCP Agent architecture, MCP protocol, and why APIs don't make good tools.
M06 — Tool Design How to design purpose-built MCP tools agents can actually use.
M07 — Advanced Workflows Skills, subagents, and hooks — the workflow composition stack.
M08 — Security Prompt injection, OWASP Top Ten for LLMs, and context rot.
M09 — Code Review Why review matters more with AI and the Writer/Reviewer pattern.
M10 — Agent Teams When parallel orchestration pays off and when it doesn't.
M11 — Post-Deployment SRE fundamentals and AI-augmented incident response.
M12 — CI/CD Integration Headless mode, pipeline patterns, and batch operations.
M13 — Team Adoption Permission modes, cost management, and the developer ownership principle.
M14 — What's Next A transferable framework for evaluating the evolving AI tool landscape.