The number on each concept is its stage. ✓ learned, ↻ changed since you learned it, "new" added since your first visit.
Claude Code study map
Get started
Install and first run 1 Common workflows 1 Best practices 1
Instructions and memory
CLAUDE.md 1 Auto memory 1 .claude directory 3 Output styles 3 Large codebases 5
Parallel work
Subagents 3 Agent view 4 Agent teams 4 Dynamic workflows 4 Git worktrees 4 Session messaging 4
Tune and debug
Models and effort 5 Prompt caching 5 Costs and usage 5 Debugging config 5
Agent SDK
SDK basics 6 Claude Code features 6 Custom tools and MCP 6 Sessions, streaming 6 Control, telemetry 6 System prompts 6 Hosting, security 6
How it works
Agentic loop 1 Built-in tools 1 Context window 1 Checkpoints, /rewind 1 Sessions 1 Slash commands 1
Where it runs
Terminal CLI 2 Terminal setup 2 VS Code, JetBrains 2 Desktop app 2 Web sessions 2 Remote and mobile 2 Slack, Claude Tag 2 Chrome, computer use 2
Automate
Headless (claude -p) 4 GitHub, GitLab CI 4 Code review 4 Routines 4 /loop and /goal 4 Channels 4 Deep links 4
Teams and enterprise
Managed settings 6 Providers and auth 6 LLM gateways 6 Network and proxies 6 Cloud environments 6 Monitoring 6
Extend Claude
Choosing extensions 3 Skills 3 Hooks 3 MCP servers 3 Plugins 3 Artifacts 4
Safety and control
Permission modes 1 Permission rules 5 Sandboxing 5 Settings scopes 5 Security and data 5
Generated from the Claude Code documentation at
code.claude.com/docs
and refreshed as it changes. Last updated September 16, 2026.