21 anatomies · One system · Endless improvement

The Claude Anatomy Library

Your complete blueprint for building, operating, and continuously improving AI-powered software teams with Claude Code.

Read the Introduction →Browse Posters

What's inside

Anatomy Docs

Themes

Posters

✓

llms.txt Ready

Foundation — The building blocks

#01

Claude Code Project Structure

The complete .claude/ layout

#02

MCP Servers

External tool connections

#03

Subagents

Specialized agents & orchestration

#04

Hook Lifecycles

Pre/Post/Stop hooks & automation

#05

CLAUDE.md

Team memory & living architecture

Reusability — Make expertise repeatable

#06

Skills

Reusable playbooks & references

#07

Plugins

Distribution patterns & security

Orchestration — Coordinate complex work

#08

Agentic Workflows

Workflow patterns & critic loops

#09

.claude/settings.json

Every key, matcher & gotcha

Workflows — Day-to-day execution

#10

Slash Commands

/command design patterns

#11

Context Management

0–80%+ usage playbook

Management — Rhythm, quality & efficiency

Prompt Engineering for Repos

System prompts & few-shots

#14

AI Cost & Token Economics

Models, batching, caching

People — Get your team productive

#15

AI Onboarding Kit

Productive in week 1

Quality — Ship with confidence

#16

AI Code Review Workflows

Gates, subagents, hooks

Operations — Run & improve

#17

AI Observability

Traces, metrics, evals, drift

#18

AI-Powered Postmortem

Blameless retros with AI

#19

AI Incident Response Playbook

Runbooks, paging, comms

Trust — Govern with care

#20

AI Governance & Risk

Policies, model cards, audits

Meta — The system that makes the library better

#21

AI Knowledge Loop

The self-improving loop