How Claude Code Works: Architecture & Internals

A technical deep-dive into Claude Code's internal mechanisms, based on official Anthropic documentation and verified community analysis.

Author: Florian BRUNIAUX | Contributions from Claude (Anthropic)

Reading time: ~25 minutes (full) | ~5 minutes (TL;DR only)

Last verified: February 2026 (Claude Code v2.1.34)

---

本篇目录

• Source Transparency
• Where Claude Code Sits in the Harness Stack
• TL;DR - 5 Bullet Summary
• Visual Overview
• Table of Contents
• 1. The Master Loop
• 2. The Tool Arsenal
• 3. Context Management Internals
• 4. Sub-Agent Architecture
• 5. Permission & Security Model
• 6. MCP Integration
• 7. Advanced Tool Use Patterns (API)
• 8. The Edit Tool: How It Actually Works
• 9. Session Persistence
• 10. Philosophy: Less Scaffolding, More Model
• 11. Claude Code vs Alternatives
• 12. Sources & References
• 13. Appendix: What We Don't Know
• Contributing
• Anthropic API Patterns for Architects
• Go further