Master Claude Code: Complete 2026 Command Guide

Claude Code is a widely adopted AI large‑model tool. The following sections enumerate its core CLI commands, describe what each command does, show the exact syntax, and list practical use cases.

1. claude

Purpose: Starts an interactive REPL session where users can converse, generate code, and analyze projects.

Typical scenarios:

Daily development Q&A

Code generation and modification

Multi‑turn debugging tasks

2. claude -p "prompt"

Purpose: Executes a single prompt in non‑interactive (Prompt) mode.

Features:

One input → one output

Does not enter interactive mode

Suitable for script‑based calls

Typical scenarios:

CI/CD automation

Quick Q&A

Shell script integration

3. claude --version

Purpose: Displays the current CLI version.

Typical scenarios:

Troubleshooting compatibility issues

Deciding whether an upgrade is needed

4. claude doctor

Purpose: Performs an environment health check.

Checks include network connectivity, API‑key configuration, local dependencies, and permission/runtime settings.

Typical scenarios:

Diagnosing launch failures

Validating environment after initialization

5. /help

Purpose: Shows command help within the interactive session.

Help content covers command list, shortcut explanations, and usage examples.

6. /init

Purpose: Initializes project context.

Usually generates or updates a project description file (e.g., CLAUDE.md) and creates a contextual understanding of the project structure.

Typical scenarios:

Onboarding a new project to Claude Code

Building a code‑understanding index

7. /clear

Purpose: Clears the current conversation context.

Effects:

Removes historical dialogue

Resets the context window

Typical scenarios:

Avoiding context pollution after long conversations

Switching tasks

8. /exit (or Ctrl + C )

Purpose: Exits the interactive Claude session.

9. /run test

Purpose: Executes a task or invokes a toolchain.

Common uses:

Running code snippets

Executing scripted tasks

Calling workflows (capabilities depend on the Claude Code integration environment)

10. /resume

Purpose: Restores a previous conversation.

Functions:

Recovers earlier dialogue context

Continues unfinished tasks

Typical scenarios:

Interrupted analysis

Ongoing large‑project iterations

11. /config

Purpose: Configures Claude Code behavior parameters.

Configurable items include default model, output style, permission policies, and API‑related parameters.

Typical scenarios:

Enterprise‑wide configuration standardization

Personalized development experience