Anthropic Releases Claude Code Terminal Agent

Anthropic recently announced Claude Code, a command-line tool that lets its Claude model act as an autonomous coding agent within developers' terminals. The tool reads and writes multi-file codebases, runs tests, manages git, and executes fixes via a think-plan-execute loop while operating locally with API-backed planning and human-in-the-loop confirmations. It targets professional engineers seeking integrated, terminal-first AI assistance.