Claude Code · Anthropic
Anthropic's terminal-native agentic coding tool and the category-defining harness, distinguished by its layered extensibility model (CLAUDE.md memory, Agent Skills, subagents, hooks, plugins, MCP) plus permission modes, checkpoints, and optional sandboxing, available from CLI/IDE/desktop/web with an Agent SDK for building on top.
Identity
| Homepage | https://claude.com/claude-code |
|---|---|
| Repository | https://github.com/anthropics/claude-code |
| License | proprietary |
| Domain | coding |
| Form factor | cli · tui · ide-extension · desktop-app · web-app · sdk · ci-step |
Model support
| Model-agnostic | partial |
|---|---|
| Native models | Claude Opus 4.x · Claude Sonnet 4.x · Claude Haiku 4.5 |
| Local models | ○ no |
| Pricing | subscription |
Extensibility
| MCP client | ✓ yes |
|---|---|
| Agent Skills support | ✓ yes |
| Plugins / hooks | ✓ yes |
Execution & safety
| Sandboxing | optional |
|---|---|
| Autonomy spectrum | supervised → semi-autonomous → autonomous-local → background-cloud |
| Subagents | ✓ yes |
| Parallel sessions | ✓ yes |
| Context management | compaction · persistent-memory |
Integration & evidence
| Headless / API | ✓ yes |
|---|---|
| GitHub integration | ✓ yes |
| Sources |
✓ profile last verified 2026-07-08 · benchmark figures are vendor-reported unless a source states otherwise