anthropics/claude-plugins-official

Official, Anthropic-managed directory of high quality Claude Code Plugins.

What it is

claude-plugins-official is a curated directory (a "marketplace") of plugins for Claude Code, Anthropic's coding assistant. It's the central place where plugins are listed so users can discover and install them.

How it's organized

The repository has two folders:

  • /plugins — plugins built and maintained by Anthropic itself.
  • /external_plugins — plugins from third-party partners and the community.

How to install

Plugins are installed through Claude Code's built-in plugin system. A user runs:

/plugin install {plugin-name}@claude-plugins-official

or browses available plugins via the /plugin > Discover menu.

Security warning (important)

The README explicitly warns that Anthropic does not control what MCP servers, files, or other software are bundled inside plugins, and cannot verify they work as intended or won't change. Users are told to trust a plugin before installing, updating, or using it, and to check each plugin's homepage for details.

What a plugin looks like

Each plugin follows a standard folder layout:

  • .claude-plugin/plugin.json — required metadata file
  • .mcp.json — optional MCP server configuration
  • commands/ — optional slash commands
  • agents/ — optional agent definitions
  • skills/ — optional skill definitions
  • README.md — documentation

Naming rules

The name field in a marketplace entry is an immutable slug. Once published, it can never change — users have it installed under that name, and renaming breaks their installs with a plugin-not-found error.

  • To change how a plugin is labeled in the UI, update displayName instead.
  • If a rename is truly unavoidable, the marketplace must include a renames map in .claude-plugin/marketplace.json (e.g., "old-name": "new-name"). The Claude Code plugin loader reads this map and automatically rewrites the old slug to the new one on the user's next sync.

Skill-bundle plugins (a special case)

When a plugin's source repository ships skills (SKILL.md files) but has no .claude-plugin/plugin.json manifest, the marketplace entry can declare the skills directly. This requires:

  • strict: false in the entry
  • an explicit skills array listing paths to skill directories

Each path is relative to the source.path and points at a folder containing a SKILL.md. Paths can go deeper than one level (e.g., ["./libA/skill-1", "./libB/skill-2"]), letting a marketplace expose a curated subset across multiple subdirectories. Each skill is registered in Claude Code as <plugin-name>:<skill-name>.

Contributing

  • Internal plugins are developed by Anthropic team members; plugins/example-plugin serves as a reference implementation.
  • External plugins can be submitted by third-party partners via a submission form. They must meet quality and security standards to be approved.

Licensing

There is no single license for the whole directory — each linked plugin carries its own LICENSE file.

More info

The README points to the official Claude Code plugin documentation for developers.

Related

  • Project
    zhaoxuya520/reverse-skill

    A cybersecurity skills router that provides AI agents with structured methodologies and tool routing for reverse engineering, penetration testing, and CTF challenges.

  • Project
    wshobson/agents

    A marketplace of Markdown‑defined plugins, agents, skills and commands that extend AI‑coding tools (Claude Code, Codex CLI, Cursor, etc.) with ready‑to‑use LLM‑driven workflows like scaffolding, testing and security scanning.

  • Project
    luongnv89/claude-howto

    A structured, example-driven learning guide that teaches developers how to master Claude Code by combining slash commands, memory, hooks, and subagents into automated workflows.

  • Project
    herdrdev/herdr

    A persistent runtime for coding agents that allows them to run in background terminals across multiple machines with session restoration and status tracking.