Claude Code
Claude Code is Anthropic's coding-agent harness. These notes describe the tool environment, not the underlying Claude model.
Summary
Claude Code is Anthropic's coding-agent harness. These notes describe the tool environment, not the underlying Claude model.
Instruction files
CLAUDE.md
Repository-level instruction files this harness reads. Treat their effective permissions as part of the setup.
Recommended
- Put durable project-specific instructions in the repository's supported
CLAUDE.mdscope. - Choose permission settings deliberately for the work being performed.
- Use non-interactive output modes only when their structured behavior fits the workflow.
Avoid
- Attributing
CLAUDE.mddiscovery or CLI permission behavior to the model. - Skipping permission checks without understanding the consequences.
Context & tool use
Claude Code exposes interactive and print modes, configurable tool permissions, and project-scoped memory behavior.
Evidence
official
- officialClaude Code supports project instruction and memory files through CLAUDE.md.
How Claude remembers your project ↗ - officialClaude Code provides interactive and non-interactive CLI modes with tool permission controls.
Claude Code CLI reference ↗