anthropics/claude-code

Claude Code is, in the project's own words, an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows, all through natural language commands. Instead of being a chat window separate from your project, it runs where you already work, so you can ask it questions about the repo you are inside, have it run shell or git commands on your behalf, or tag @claude on GitHub. The README points users to the official documentation for the deeper guide and shows the basic install paths: a one-line shell installer for macOS and Linux, a Homebrew cask, a PowerShell installer for Windows, a WinGet package, and an older npm install that the README marks as deprecated. After installing, you navigate to your project directory and run the claude command. The repository also bundles a plugins directory that extends the tool with extra commands and agents, with separate documentation linked from the main README. The README mentions a /bug command for reporting issues, a Claude Developers Discord for community support, and a section about data collection: feedback such as code acceptance or rejection and chats from sessions are retained under stated retention limits, and the README explicitly says feedback is not used for model training, with full details deferred to the commercial terms and privacy policy. You would use Claude Code when you want an AI assistant that can read and act on your local code rather than just answer in a separate chat. The repository's primary listed language is Shell, reflecting the install scripts and command wrappers shipped here.