Zero to operator · 10 chapters
Claude Code: From First Session to Reliable Engineering
A plain-language, hands-on path from opening a project to operating safe, verifiable agentic workflows.
Learn how Claude Code sees a project, how to give it useful context and bounded authority, how to review and verify changes, and how to scale from one supervised task to repeatable team workflows.
Begin chapter oneDeliberate progression
The curriculum
Foundation chapters give you the map. Builder chapters teach you to make controlled changes. Operator chapters teach permissions, recovery, quality, and proof.
- 01FoundationKnow What You Are OperatingInstall Claude Code, choose the right surface, and complete a first supervised loop without confusing the assistant, the repository, or the running application.14 min lesson + 45 min lab
- 02FoundationOrient to an Existing ProjectTurn an unfamiliar repository into a usable system map before allowing an agent to implement changes.16 min lesson + 45 min lab
- 03FoundationTurn Intent Into a Build SpecificationTranslate business judgment into a bounded task packet Claude Code can implement and verify without guessing at the important decisions.18 min lesson + 60 min lab
- 04BuilderMake Project Context DurableUse CLAUDE.md, rules, skills, and memory deliberately so each session begins with useful truth instead of an ever-growing instruction dump.18 min lesson + 45 min lab
- 05BuilderBound Permissions and the SandboxReplace approval fatigue with an explicit authority model that combines permission rules, execution modes, sandbox boundaries, and human judgment.18 min lesson + 60 min lab
- 06BuilderMake Changes With Git and WorktreesUse Git history, branches, diffs, checkpoints, and worktrees to keep agent-generated changes inspectable, isolated, and recoverable.18 min lesson + 60 min lab
- 07BuilderDebug and Verify the ResultUse Claude Code to reproduce failures, test hypotheses, and close the loop with evidence instead of accepting a plausible patch.18 min lesson + 60 min lab
- 08OperatorExtend With MCP, Skills, and HooksChoose the right extension mechanism for external tools, reusable knowledge, and deterministic automation without turning every capability into permanent context or authority.20 min lesson + project lab
- 09OperatorDelegate and Run Work in ParallelUse subagents, parallel sessions, agent teams, and worktree isolation only when task boundaries and integration ownership make concurrency safer and faster.20 min lesson + project lab
- 10OperatorOperate in CI and Ship the CapstoneTurn one proven interactive workflow into bounded automation with machine-readable output, CI controls, operating receipts, and a clear path back to human review.22 min lesson + capstone
Capstone
Ship one trustworthy Claude Code workflow
Choose a real workflow you understand, use Claude Code to implement one complete vertical slice, and leave behind enough project context, controls, tests, and evidence that another person can safely repeat the work.
What you must build
- ◆A written problem statement, user outcome, scope boundary, and explicit non-goals.
- ◆A repository map that identifies the canonical checkout, main runtime, source of truth, and verification commands.
- ◆A concise project CLAUDE.md containing only durable commands, conventions, architecture facts, and safety rules.
- ◆A permission and sandbox policy that protects secrets and requires approval for consequential or external actions.
- ◆An isolated Git branch or worktree with a reviewable diff and no unrelated changes.
- ◆A test or observable check that fails before the repair when practical and passes afterward.
- ◆A runbook for repeating the workflow interactively or in CI with bounded tools, turns, cost, and stopping conditions.
How you prove it
- The final diff contains only the intended scope and can be explained file by file.
- The relevant test, typecheck, lint, build, or visual verification succeeds in the named environment.
- No secret value appears in source, prompts, logs, screenshots, artifacts, or Git history.
- A second reviewer or verification pass checks the result against the original acceptance criteria.
- The workflow documents rollback, escalation, and the evidence required before any future increase in autonomy.
