Claude Code
Anthropic's official CLI that lets Claude read, write, and execute code in your terminal — the most capable way to use Claude for software engineering.
Claude Code is a Node.js command-line interface that runs Claude as a coding agent in your project. Unlike chat interfaces, it has direct access to your filesystem, git, the shell, your editor and (via MCP) any external service you wire up. Sessions remember the project's CLAUDE.md and any per-project skills, agents, hooks and MCP server configurations.
It's powered by Claude Opus 4.7 (with optional 1M-token context window for huge codebases), Claude Sonnet 4.6, or Claude Haiku 4.5 — pick the model per session or even per subagent. Tool support includes Read, Edit, Write, Bash, Grep, WebFetch, plus any MCP server you register.
Common workflow: launch claude in your repo, describe the task, watch it explore the codebase, propose a plan, edit files, run tests, and commit. Hooks (PreToolUse, PostToolUse, Stop) let you guardrail or automate around its actions.