Loop
Dashboard

Tracks

Data Structures & AlgorithmsSystem DesignInterview PlaybookAI Engineering with Claude
ProblemsProgress
0-day streak
Loop
  1. AI Engineering with Claude
  2. Claude Code Mastery

Claude Code Mastery

The full toolkit: memory, modes, hooks, skills, subagents, and session hygiene.

0 of 7 lessons complete · 93 min total

Start module
  1. CLAUDE.md: Persistent Project MemoryWhat belongs in project memory (commands, gotchas, style rules), what doesn't, and keeping it true as an operating manual.Lesson · 14 minNot completed
  2. Plan Mode & Permission ModesRead-only planning before edits, permission levels, and matching mode to task risk.Lesson · 14 minNot completed
  3. Hooks: Deterministic AutomationLifecycle-event hooks for guaranteed behaviors — formatters, gates, notifications — that prompting alone can't promise.Lesson · 15 minNot completed
  4. Skills & Slash CommandsPackaging on-demand expertise into reusable skills and when a slash command beats re-explaining a workflow.Lesson · 14 minNot completed
  5. Custom SubagentsDefining specialized subagents with scoped tools and prompts, and when delegation beats doing it in the main thread.Lesson · 14 minNot completed
  6. Session ManagementCourse-correcting early, clearing often, and recognizing when a conversation has rotted past saving.Lesson · 13 minNot completed
  7. Claude Code CheckpointFeature selection: memory vs hooks vs skills vs subagents for a given need.Checkpoint · 9 minNot completed
Previous modulePrompting FoundationsNext moduleAgent Patterns

© 2026 Kitchen Labs

PrivacySupport