Hand-building every instruction, agent or connection so your AI assistant handles a recurring task well eats hours. This is Anthropic's official catalog of bundles (plugins) that install into Claude Code with one command. Each bundle can ship skills, agents, hooks and MCP servers.
Who this is NOT for
If you use Claude in the claude.ai chat or in Cowork rather than the terminal, this catalog isn't for you: the docs themselves send you to a different page. You don't need it either if you only want a single skill, since a skill saved in ~/.claude/skills/ works without a plugin. And if you can't review what you install, leave /external_plugins alone: those are third-party plugins, and Anthropic says it doesn't verify what they do.
"Official" doesn't mean safe. The README separates /plugins, maintained by Anthropic, from /external_plugins, which comes from partners and the community. For the external ones, Anthropic admits it doesn't control the MCP servers or software they ship, and that they can change. Almost everyone makes the same mistake: installing ten plugins "just in case". Each one runs with your permissions and puts its description in the context of every session. With over 1,000 open issues and PRs and commits from yesterday, the project is alive and moves fast. Install one, measure it, and only then add the next.
Your first step
Open Claude Code, type /plugin and go to the Discover tab. Pick one Anthropic-built plugin, check its context cost estimate, and only then install it with /plugin install {name}@claude-plugins-official.
The Claude Code Plugins Directory implementation guide
The project belongs to Anthropic (anthropics) and it is free: the link beside asks for nothing. The guide to set it up without getting lost is ours: step by step, common mistakes and a checklist. Leave your email and we will send it.
No spam. Unsubscribe anytime.