Type: url
Original: https://code.claude.com/docs/en/how-claude-code-works
Status: generated
Summary
Anthropic’s Claude Code documentation page describing the tool’s core architecture: an agentic loop of gathering context, taking action and verifying results, powered by models that reason and tools that act, with Claude Code itself being the surrounding agentic harness. It covers what Claude Code can access, its execution environments and interfaces, how sessions are stored, resumed and forked, how context is managed and compacted, checkpoints and permission modes as safety mechanisms, and tips for working with it.