Claude Code: Anthropic's coding agent
Claude Code is Anthropic's agentic coding tool. It reads your codebase, edits files, runs commands and works with your development tools, and it runs in the terminal, in VS Code and JetBrains IDEs, in the Claude desktop app and in the browser. Anthropic made it generally available on May 22, 2025.
Claude Code is for developers and for people learning to build software with an AI agent. Where a personal agent works across your apps and accounts, Claude Code works inside a code repository: it plans a change, edits the files, runs your tests and can commit the result. Most surfaces run on a Claude subscription or an Anthropic Console account, and the terminal CLI, VS Code and JetBrains can also run on a third-party model provider. The desktop app is the one surface the docs say requires a paid subscription. New to it? Start with the Claude Code beginner's guide.
#How do you install Claude Code?
On macOS, Linux and WSL the native installer is one command:
curl -fsSL https://claude.ai/install.sh | bashOn Windows, run irm https://claude.ai/install.ps1 | iex in PowerShell, run the install.cmd script in CMD, or install with WinGet (winget install Anthropic.ClaudeCode). On macOS you can also use Homebrew (brew install --cask claude-code). Then run claude inside a project folder and log in when it asks.
#How do Claude Code skills work?
A skill is a folder with a SKILL.md file: optional YAML frontmatter, then the instructions Claude follows. Personal skills live in ~/.claude/skills/ and project skills in .claude/skills/, one folder per skill. The folder name becomes the command, so .claude/skills/deploy/SKILL.md runs as /deploy, and Claude can also start a skill on its own when the skill's description matches the task. Claude Code extends the Agent Skills open standard with fields of its own, such as disable-model-invocation and context.
For a worked example of writing a skill, see custom skills and slash commands; for running your own checks around them, hooks for beginners.
Key facts
| Maker | Anthropic |
|---|---|
| Type | Coding agents |
| Status | Generally available |
| Price | Included with Claude Pro ($20 a month, $17 billed annually) and Max (from $100 a month); or Anthropic API billing |
| Availability | Terminal on macOS, Linux, WSL and Windows; VS Code and JetBrains; desktop app; web and the Claude mobile app |
| Your own skills | Yes. Agent Skills SKILL.md in .claude/skills, plus plugins that bundle skills, agents, hooks and MCP servers (skills docs) |
| Website | code.claude.com |
| Docs | Claude Code docs |
| Last verified |
Want a guided start? Claude Code for Builders, the ClawDocx playbook
Start here
- Multi-Agent Workflows
An advanced guide to running multiple OpenClaw agents: per-agent workspaces, agent-to-agent communication, and integrating multi-agent setups with Claude Code.
Skills
- How to build a Claude Code plugin and share it through a marketplace
Build a Claude Code plugin from one directory, check it with claude plugin validate, then list it in a marketplace.json so your team installs it by name.
- How to use your Claude Code skills in Cursor
Cursor loads SKILL.md folders from .claude/skills for compatibility, so your Claude Code skills work unmoved. Fix the name field, then check what carries.
- Cursor vs Claude Code vs Copilot (2026): Agents, Skills and Pricing
How Cursor, Claude Code and GitHub Copilot compare in 2026 on agent surfaces, skills folders, model access and the prices each vendor publishes.
- Codex vs Claude Code (2026): pricing, limits, skills and config
Codex and Claude Code compared on documented plans, usage limits, where skills live and which config files each one reads, with a source for every figure.
- Claude Code skills vs subagents vs plugins vs MCP servers: what to use when
Skills, subagents, plugins and MCP servers solve four different problems in Claude Code. Here is where each one lives, what loads it, and how to choose.
- Claude Code Skills: Build Custom Slash Commands That Actually Work (2026 Guide)
The complete guide to creating Claude Code skills: custom slash commands powered by SKILL.md files, with build, share, and chain steps.
- Claude Code in 2026: Voice Mode, /loop, Scheduled Tasks, and Everything New
A practical guide to Claude Code's biggest 2026 features: voice mode, /loop for recurring tasks, scheduled tasks, auto-memory, Skills 2.0, and worktrees.
Latest news
- Claude Code 2.1.302: the three default changes to check before you upgrade
Claude Code v2.1.302 was tagged on 16 October 2026. Three entries change defaults for scripted, desktop and VS Code sessions rather than adding features.
- Claude Code 2.1.319 changes auto mode when telemetry is off, and claude -p now exits 1
Claude Code 2.1.319 stops auto mode falling back to its built-in classifier when telemetry or feature flags are off, and makes claude -p exit 1 on refusal.
- Claude Code 2.1.299 tightens MCP allow rules, MCP redirects and skill file names
Claude Code 2.1.299 shipped on 14 October 2026: MCP allow rules now reach only the server they name, and MCP requests stop following redirects.
- A tampered Claude Code extension on Cursor's marketplace installed an SSH backdoor
Cursor says tampered Claude Code extension builds 2.1.312 and 2.1.313 reached its marketplace on 29 October. Here is how to check whether you are affected.
- Claude Code 2.1.305 and 2.1.306: the MCP allowlist change to check before you upgrade
Claude Code 2.1.305 stops filtering some managed MCP servers through allowedMcpServers, adds a sandbox namespace switch and fixes a 2.1.304 regression.
- Claude Code 2.1.316: fast mode on Opus 4.8 is deprecated, with removal on 30 November 2026
Claude Code 2.1.316 deprecates fast mode on Claude Opus 4.8 and sets removal for 30 November 2026, after which those requests return an error.
- Claude Code 2.1.309 and 2.1.310: the sandbox and permission fixes to take now
Claude Code 2.1.309 fixes the sandbox, Bash permission checks, deny rules and hooks on macOS, Linux and Windows. 2.1.310 adds the Monitor tool.
- Claude Code 2.1.304: Code Review levels, mod controls and skill hook caps
Claude Code 2.1.304 adds Quick, Standard and Deep levels to Code Review, two new mod tool.check fields, and per-source caps on skill hooks.
- Claude API OpenTelemetry tracing: what the beta exports and how to turn it on
Anthropic launched OpenTelemetry export for the Claude API in beta on 14 October 2026. What the spans carry, the two setup paths, and the limits.
- Anthropic's 2026 Usage Policy update: what changes for people building agents
Anthropic published its 2026 Usage Policy on 8 October, effective 12 November. The sections that change what an agent may do, and what stays the same.
- Claude Code now runs mods: what they change and what they can reach
Anthropic shipped mods in Claude Code v2.1.287 on 1 October 2026. They run inside the agent with your permissions, unsandboxed, and here is what they reach.
- Anthropic's Very Bad Week: Claude Code Source Leak, a Secret Model and a Supply Chain Attack
In one week Anthropic leaked Claude Code's full source on npm, exposed details of an unreleased model called Mythos, and npm suffered a supply chain attack.
- Claude Code Channels & Auto Mode: Is Anthropic Coming for OpenClaw?
Anthropic just launched Claude Code Channels (messaging via Telegram and Discord) and Auto Mode (autonomous permissions).
- Claude Code Just Got Dispatch & Channels — Here's How to Build Apps 10x Faster in 2026
Claude Code added Dispatch, which controls coding agents from your phone, and Channels, which connects them to Telegram and Discord. Here's how to use both.
- The OpenClaw Clone Wars: Every AI Agent Framework Competing for Your Computer in 2026
OpenClaw started the autonomous AI agent wave. Now competitors range from Claude Cowork to minimalist tools to enterprise platforms. Here's how they compare.
Comparisons
- AGENTS.md vs CLAUDE.md: how Codex and Claude Code read project instructions
Write one AGENTS.md. Claude Code reads it when no CLAUDE.md sits in your working directory or above, and Codex reads nothing else. Here are the exact rules.
- Getting Started with Claude Code + OpenClaw: The Complete Setup Guide
Learn how to connect Claude Code to OpenClaw using ACP: run coding sessions from chat, spawn thread-bound agents, and use voice mode, /loop, and hooks.
- I Gave My AI Agent $100 and 7 Days to Build a Business — Here's What Happened
What happens when you give an OpenClaw AI agent a $100 budget and one week to build a real online business from scratch? We ran the experiment.
- How to Vibe Code a Profitable App with OpenClaw + Claude Code (The 2026 Playbook)
Vibe coding is MIT's breakthrough tech of 2026. Here's how to use OpenClaw and Claude Code to go from idea to profitable app without writing code yourself.
More articles
- Claude Code pricing: how much do Pro, Max, Team and the API cost?
What Claude Code costs on every Claude plan: Free, Pro at $17 or $20, Max 5x and 20x, Team seats and Enterprise, plus the per-token API rates.
- Claude Code April 2026: Computer Use, Auto Mode, Cloud Execution, and the Shift to Persistent Agents
Claude Code's biggest updates yet: computer use for GUI automation, auto mode for fewer prompts, scheduled cloud tasks, and remote control from your phone.
- Best Claude Code Settings & Configuration (2026): Power User Guide
Optimize your Claude Code setup with the best settings for CLAUDE.md, hooks, model routing, effort levels, permissions, and sub-agent configuration.
- Claude Code for Beginners: Build Your First App in 30 Minutes (2026)
A step-by-step beginner's guide to Claude Code: install it, configure your project, learn the essential commands, and build your first app in under 30 minutes.
- The Ultimate Claude Code Guide (2026): Everything You Need to Know
The complete guide to Claude Code in 2026, from installation and essential commands to advanced patterns for real engineering workflows.
- Claude Code Agent Teams: Parallel Development with Multiple AI Instances
A guide to Claude Code's experimental Agent Teams feature: coordinate multiple Claude instances on shared tasks with messaging and centralized management.
- Claude Code Power User: /effort, /context, /loop, and the Slash Commands That Actually Matter
A practical guide to Claude Code's most useful slash commands: /effort for analysis depth, /context for memory issues, /loop for recurring tasks, and /simplify.
- Claude Code Hooks: Automate Your Workflow Without Leaving the Terminal
A beginner-friendly guide to Claude Code hooks — the automation system that runs shell commands, HTTP calls, and LLM prompts at key lifecycle points.
- Claude Code Memory: CLAUDE.md, Auto-Memory, and Making Claude Remember Your Project
A practical guide to Claude Code's memory systems — CLAUDE.md instruction files, auto-memory, .claude/rules, and memory management.
- Claude Code Sub-Agents: Delegate, Specialize, and Scale Your Coding Workflows
Learn Claude Code's sub-agent system: built-in agents like Explore and Plan, custom specialists, routing tasks to cheaper models, and scaling complex projects.
Frequently asked questions
- Is Claude Code free?
- No. Claude Code is not part of the free Claude plan. It comes with Pro ($20 a month, or $17 a month billed annually), Max (from $100 a month), Team and Enterprise, and it also works with an Anthropic Console account billed for API use.
- On Windows, is it better to install natively or in WSL?
- Both work. A native install needs Git for Windows if you want Claude Code's Bash tool; without it, Claude Code uses PowerShell as the shell tool instead. In WSL you get the same one-line installer as macOS and Linux and a Bash shell, and the docs note that WSL setups do not need Git for Windows.
- What is the difference between a Claude Code skill and a plugin?
- A skill is one folder of instructions with a SKILL.md file. A plugin is a package that can bundle several skills with agents, hooks and MCP servers. Adding a .claude-plugin/plugin.json file turns a skill folder into a plugin, and plugin skills run as /plugin-name:skill-name.