Claude Code is an agentic coding tool that runs in your terminal. It reads your codebase and carries out work you describe in natural language: executing routine tasks, explaining code, and handling git workflows. You can use it from the terminal, from an IDE, or by tagging @claude on GitHub.
The repository also ships plugins that extend Claude Code with custom commands and agents, and a /bug command reports issues from inside the tool. Usage data such as code acceptance or rejection, associated conversation data, and feedback submitted through /bug are collected under the published data-usage policy, with limited retention for sensitive information and a policy against using feedback for model training.
Features
- Natural-language commands: describe a task and the tool plans and executes it against your codebase
- Codebase understanding: reads project files to answer questions and explain code
- Git workflows: handles branching, commits, and related version-control steps as part of a task
- Multiple surfaces: terminal, IDE integrations, and GitHub via @claude mentions
- Plugins: bundled plugins add custom commands and agents
- In-tool bug reporting: the /bug command files feedback directly
Extensions & skills
Agency Agents
Collection of specialized AI agent personas that install into Claude Code, Cursor, Codex, and other coding assistants
Agenta
Workspace for building agents through chat and running them in the background, with swappable harnesses and models
Agentlas OS
Agent OS: keep specialist agents in a hub, spin up a temporary orchestrator per task. Local-first, works with any model.
Claude-Mem
Compresses Claude Code sessions into a local SQLite and Chroma memory store, queryable from other agents over MCP
Built with this
Claude Agent SDK
SDK that exposes the Claude Code agent loop for building custom agents.
Open Cowork
Open-source AI agent desktop app for Windows & macOS. One-click install Claude Code, MCP tools, and Skills — with sandbox isolation, multi-m
Remote Factory
Domain-agnostic multi-agent software design and evolution harness
Shepherd
Records agent runs as reversible, inspectable execution traces whose outputs are reviewed before they are applied or discarded
Integrated by
Agent Skills
Portable skill bundles teaching AI coding assistants WordPress block, theme, plugin, and REST API development patterns
Agent Vault
HTTP credential proxy that injects secrets into agents' outbound requests so agents never hold real credentials
Agenvoy
Make AI actually work for you - A personal agent that writes its own tools and repairs itself.
Apify MCP Server
MCP server exposing Apify Store Actors as dynamically discovered tools for agents.
Alternatives
Ante Preview
Ghost in your shell. Ante is a self-contained agent harness with a highly optimized core. It works like Claude Code or Codex, with none of t
Caveman Code
Terminal coding agent that answers in a compressed caveman register to cut per-turn output tokens
DeterminFlow
A production-oriented AI workflow runtime for building, validating, recovering, and shipping complex AI workflows as dependable services. 面向
FuXi
FuXi is a fast, self-contained AI coding agent that lives in your terminal — edit code, run commands, and drive tools, with cost-aware routi
Mentioned with
Agent Reach Lower-confidence signal
CLI that lets AI agents read and search Twitter, Reddit, YouTube, GitHub, Bilibili, Xiaohongshu, RSS, and the web
LiteLLM Agent Platform Lower-confidence signal
Self-hosted control plane giving one API and UI over multiple agent runtimes, with sessions, cron schedules and memory
-
WRUD Lower-confidence signal
Open-source CLI that logs each agent session's tools, files and tokens to local SQLite and flags where a large model did trivial work