This is the overview page for Claude Code. See also: Setup & IDE Integration, Context & Memory, Skills & Plugins.

Claude Code

Claude Code is Anthropic’s official CLI and VS Code extension for AI-powered software development. It gives Claude direct access to your codebase — reading files, writing code, running commands, and managing Git — all from your terminal or IDE.

What Makes Claude Code Different

Unlike chat-based AI assistants, Claude Code operates inside your development environment with full filesystem access. It doesn’t just suggest code — it reads your entire project, understands the architecture, makes targeted changes, runs tests, and commits to Git.

CapabilityDescription
Codebase awarenessReads and indexes your entire project structure
File operationsCreates, edits, and deletes files directly
Command executionRuns shell commands (build, test, deploy)
Git integrationCommits, branches, creates PRs
MCP connectionsConnects to databases, APIs, external tools
Multi-sessionParallel instances for different features

Key Concepts at a Glance

Models Available

ModelIDBest For
Opus 4.6claude-opus-4-6Complex tasks, architecture, deep reasoning
Sonnet 4.6claude-sonnet-4-6Routine development, balanced speed/quality
Haiku 4.5claude-haiku-4-5-20251001Quick tasks, trivial operations
Opus 4.6 is more expensive but solves problems correctly the first time, often saving tokens overall. Use Sonnet for routine work and Haiku for trivial tasks.

The PSB Workflow — Plan, Setup, Build

This structured methodology separates effective Claude Code users from beginners.

1

1. PLAN — Before Any Code

Define your MVP clearly. Create a project spec doc covering product requirements and engineering requirements. Use Plan Mode to design architecture without writing code. Ask Claude to interview you about technical decisions.
2

2. SETUP — Environment Checklist

Create GitHub repo, .env with API keys, CLAUDE.md with project rules, docs/ folder structure, .claude/settings.json with permissions, install relevant MCPs and skills.
3

3. BUILD — Feature by Feature

Research, Plan, Implement, Test, Next feature. Build one feature at a time, write tests, verify they pass, then move on. Use GitHub Issues as your source of truth — tag Claude in issues and it creates PRs automatically.

Permission Modes

ModeBehaviorToggle
DefaultAsks permission for every file change or commandShift+Tab
Accept EditsEdits files autonomously, may still ask for commandsShift+Tab
Plan ModeRead-only — thinks and plans, writes zero codeShift+Tab
Bypass PermissionsAuto-accepts ALL tool calls (YOLO mode)--dangerously-skip-permissions flag

Essential Commands

CommandPurpose
/modelSwitch between Opus, Sonnet, Haiku
/initScan codebase and generate CLAUDE.md
/memoryEdit persistent rules
/compactCompress conversation to save tokens
/clearFresh start for unrelated tasks
/contextVisualize token usage
/mcpManage MCP server connections
/fastToggle fast output mode

The 10 Mantras

  1. Read first, answer second — Never guess when you can verify
  2. Simplicity is sophistication — The best code is the least code
  3. Make it work, make it right, make it fast — In that order
  4. Convention over configuration — Follow established patterns
  5. Fail fast, fail loud — Errors should be obvious, not silent
  6. Measure, don’t assume — Profile before optimizing
  7. Security is not optional — Build it in from the start
  8. Documentation is code — Keep them in sync
  9. User success is my success — Focus on outcomes
  10. Excellence is a habit — Consistent quality, not occasional brilliance