English | 中文 | 日本語

Learn Claude Code

A teaching repository for implementers who want to build a high-completion coding-agent harness from scratch.

This repo does not try to mirror every product detail from a production codebase. It focuses on the mechanisms that actually decide whether an agent can work well:

• the loop
• tools
• planning
• delegation
• context control
• permissions
• hooks
• memory
• prompt assembly
• tasks
• teams
• isolated execution lanes
• external capability routing

The goal is simple:

understand the real design backbone well enough that you can rebuild it yourself.

What This Repo Is Really Teaching

One sentence first:

The model does the reasoning. The harness gives the model a working environment.

That working environment is made of a few cooperating parts:

• Agent Loop: ask the model, run tools, append results, continue
• Tools: the agent's hands
• Planning: a small structure that keeps multi-step work from drifting
• Context Management: keep the active context small and coherent
• Permissions: do not let model intent turn into unsafe execution directly
• Hooks: extend behavior around the loop without rewriting the loop
• Memory: keep only durable facts that should survive sessions
• Prompt Construction: assemble the model input from stable rules and runtime state
• Tasks / Teams / Worktree / MCP: grow the single-agent core into a larger working platform

This is the teaching promise of the repo:

• teach the mainline in a clean order
• explain unfamiliar concepts before relying on them
• stay close to real system structure
• avoid drowning the learner in irrelevant product details

What This Repo Deliberately Does Not Teach

This repo is not trying to preserve every detail that may exist in a real production system.

If a detail is not central to the agent's core operating model, it should not dominate the teaching line. That includes things like:

• packaging and release mechanics
• cross-platform compatibility layers
• enterprise policy glue
• telemetry and account wiring
• historical compatibility branches
• product-specific naming accidents

Those details may matter in production. They do not belong at the center of a 0-to-1 teaching path.

Who This Is For

The assumed reader:

• knows basic Python
• understands functions, classes, lists, and dictionaries
• may be completely new to agent systems

So the repo tries to keep a few strong teaching rules:

• explain a concept before using it
• keep one concept fully explained in one main place
• start from "what it is", then "why it exists", then "how to implement it"
• avoid forcing beginners to assemble the system from scattered fragments

Recommended Reading Order

The English docs are intended to stand on their own. The chapter order, bridge docs, and mechanism map are aligned across locales, so you can stay inside one language while following the main learning path.