Claude Code is an agentic AI coding tool developed by Anthropic that lives directly in your terminal, understands your entire codebase, and helps you code faster through natural language commands. It can read and analyze code, edit files across your project, run terminal commands, and manage Git workflows autonomously, making it a powerful AI pair programmer that operates from the command line. Claude Code addresses the need for an AI assistant that deeply integrates with existing development workflows rather than requiring developers to switch to a separate IDE or web interface.
Claude Code features extended thinking for reasoning through complex problems before writing code, a checkpoint system that automatically saves code state before each change with instant rewind capability, and deep integrations with GitHub and GitLab for reading issues, writing code, running tests, and submitting pull requests. It supports project memory through CLAUDE.md files, task queuing for complex multi-step operations, and configurable permission controls for automated workflows. Claude Code works with Anthropic's current models, Claude Opus 5.5, Sonnet 5.5 and Haiku 5.5, plus Claude Fable 5.1 where it is available, selectable through aliases such as opus, sonnet, haiku and fable, providing flexibility between capability and speed.
Claude Code is designed for professional developers who want AI assistance without leaving their terminal-based workflow. It is available as a CLI tool, as native extensions for VS Code, Cursor, Windsurf, and JetBrains IDEs, and through a desktop application. Claude Code excels at large-scale refactoring, codebase exploration, test generation, and handling routine development tasks, making it particularly valuable for developers working on complex projects who want to maintain their existing Git-based development practices.
Code Review for Claude Code is generally available on Team and Enterprise plans. After an organization Owner turns it on and installs the Claude GitHub App, several agents review each GitHub pull request alongside the surrounding code, a verification step filters out false positives, and findings are posted as inline comments ranked by severity, with a summary and a check run; a review never approves or blocks a merge. Owners choose a Quick, Standard or Deep review level per organization or repository and when reviews run. Reviews are billed on token usage through usage credits, separately from plan usage, with no flat per-review price and a monthly spend limit set by the organization. On any plan, /code-review runs a lighter review locally in a Claude Code session.
Claude Code's /design-sync and /design-login commands now show a notice that the standalone Claude Design app (claude.ai/design) closes by December 14, 2026.
Claude Code can be extended with plugins that bundle skills, subagents, hooks and MCP servers into one installable unit. Anthropic maintains an official plugin marketplace, claude-plugins-official, which Claude Code adds the first time you start an interactive terminal session unless a managed policy blocks it. You browse and install plugins from the /plugin panel, and teams can add their own marketplaces or load a plugin straight from a local folder.
Security note (October 2026): On October 29, 2026, tampered copies of the Claude Code extension (versions 2.1.312 and 2.1.313) were distributed for several hours through Cursor's extension marketplace only. Microsoft reported that the packages were never published on Microsoft's Visual Studio Marketplace or npm, and that Anthropic confirmed its systems were not compromised. Cursor users who installed or updated the extension from its marketplace on that day should follow Cursor's security notice.
Message threads (public beta, November 2026) are on by default for most Claude Code users; Anthropic reports about 35% lower median time to first token on 500K–1M-token requests, with token usage, pricing and rate limits unchanged.




