Claude Code

What is Claude Code

Claude Code is Anthropic’s flagship intelligent coding tool for the terminal, capable of quickly converting text descriptions into code.

How to Install

For detailed installation instructions, please refer to the official documentation

Common Commands

Command	Function	Example
claude	Start interactive mode	claude
claude “task”	Execute a one-time task	claude “fix the build error”
claude -p “query”	Execute a one-time query and exit upon completion	claude -p “explain this function”
claude -c	Continue the most recent conversation in the current directory	claude -c
claude -r	Resume a previous conversation	claude -r
claude commit	Create a Git commit	claude commit
/clear	Clear conversation history	> /clear
/help	Display available commands	> /help
exit or Ctrl+C	Exit Claude Code	> exit

Working Modes

Anthropic’s Claude Code has two working modes:

1. Execute / Normal Mode: Directly converts user prompt requirements into code.
2. Plan Mode: A “plan first, code later” constraint mode. Its core value lies in controllability, reviewability, and reducing risks for complex tasks.

You can think of Normal Mode as a working engineer, while Plan Mode is an architect who draws diagrams first.

You can switch modes during runtime by pressing Shift + Tab, or enable Plan Mode by default using the following command: claude --permission-mode plan

Set Plan Mode as default:

// .claude/settings.json
{
  "permissions": {
    "defaultMode": "plan"
  }
}

Agents

Agents in Claude Code are equivalent to a collection of intelligent entities that can accomplish complex goals through “multiple role-based, memory-enabled, collaborative” Claude sub-instances. Their core value can be summarized as: Division of Labor + Collaboration + Stability.

Each Agent has its own context that won’t interfere with others, and can be repeatedly invoked for continuous work.