How Claude Code works
Each flow is an ordered chain of real prompts from the catalog: system sections, tools, and agents, showing how they reinforce each other in typical tasks. Open a flow for an overview diagram and a step-by-step breakdown with links to every prompt.
Fixing a Bug
From symptom to patch using search, read, edit, and verification
Exploring a Codebase
Discovery with glob, grep, reads, and optional Explore agent
Project Onboarding
How /init bootstraps project understanding and generates CLAUDE.md
Planning & Task Management
Plan mode, Plan agent, todos, and background tasks
Autonomous Mode
Self-driving coding with periodic critique and course-correction
Multi-Agent Orchestration
Coordinator mode, Agent tool, and worker personas
Swarm & Team Collaboration
Coordinated multi-agent teams with shared state and inter-agent messaging
Memory & Context
Environment injection, memory taxonomy, filtering, extraction, and output shaping
Context Compaction
How growing conversations are compressed to stay within context limits
Skills & Self-Improvement
How Claude Code discovers, executes, and iteratively refines reusable skills
Hooks & Automation
Shell hooks, scheduled tasks, and command safety policies
MCP & External Tools
Extending capabilities with Model Context Protocol servers and resources
Session Insights
The /insights analytics pipeline that turns usage history into a personalized report
Web Research & Integration
Search, fetch, then write and validate in the repo
Flows are editorial paths for learning; the live product composes prompts dynamically via source-level assembly.