Master Claude Code: Complete Guide to Every cc Command for AI Agent Development

Yesterday I planned to write an article, but first I created “Understanding Agent Architecture to See Through Claude Code CLI Options” for those unfamiliar with agent core components; today I systematically present all cc commands involved in a typical R&D workflow.

Configuration Process Overview

Self‑help menu (self‑rescue and self‑check)

Appearance configuration and editor integration

Core capability component configuration (LLM + Action + Prompt + Tool + Memory)

Multi‑person development workflow (Git and GitHub) common settings

Conversation export

Self‑Help Menu

Use /help to display help and /status to check the current state.

IDE Integration and Appearance

Connect cc to your editor with /ide, which synchronizes selections between the editor and the terminal.

Configure the theme and other UI aspects with /config.

Add a custom status line using /statusline.

Core Capability Configuration

Switch the underlying large language model with /model.

Set permissible actions via /permissions.

Define the agent’s role and prompts with /memory.

Install additional tools, e.g., Playwright MCP, using /mcp and the command:

claude mcp add shared-server --scope project npx @playwright/mcp@latest

Inspect the current memory context with /context.

Local Development Commands

Switch the model with /model, clear the context with /clear, compress it with /compact, and add a backend API directory with /add-dir.

Collaboration Commands

View PR comments using /pr-comments.

Review a PR with /review.

Run a security review of the current changes with /security-review.

Conversation Export

Export the dialogue for later review or sharing with /export.

Summary

The article provides a complete, chronologically ordered guide to Claude Code’s command‑line interface, helping developers avoid memorizing scattered commands and instead understand the tool’s overall architecture and workflow.