Free Tool

Free Claude Code Plugins Directory

A Claude Code plugins directory helps developers find reusable extension patterns for coding agents. Search MCP servers, slash commands, subagents, hooks, settings presets, output styles, and GitHub automation ideas, then copy a setup prompt with a concrete verification check.

Search Claude Code extensions

Filter by extension type, agent fit, or workflow. Pick a result to generate a setup prompt and verification check.

15 matches

How to choose a Claude Code plugin

  1. 1Search by workflow. Enter a use case such as GitHub, browser QA, security review, docs, or test fixing.
  2. 2Filter by extension type. Choose MCP server, slash command, subagent, hook, settings, GitHub Action, or output style.
  3. 3Review the setup card. Check the suggested install location, agent fit, complexity, and acceptance check.
  4. 4Copy the setup prompt. Paste the generated prompt into Claude Code or another coding agent and keep the verification step attached.
  5. 5Start with least privilege. Grant the smallest useful scope first, then expand access only after the plugin works in a real project.

Plugin patterns worth separating

Capabilities

MCP servers and GitHub Actions connect agents to external systems. Treat them as capability plugins and scope tokens carefully.

Behavior

Slash commands, output styles, and project settings shape how Claude Code plans, reviews, explains, and validates work.

Guardrails

Hooks and focused subagents catch risky edits, run repeatable checks, and keep review attention on a narrow domain.

Claude Code plugins FAQ

What is a Claude Code plugin?

A Claude Code plugin is a reusable extension pattern for Claude Code, such as an MCP server, slash command, subagent, hook, settings preset, output style, or GitHub workflow that adds a repeatable capability to an agent workflow.

What should I install first for Claude Code?

Start with project settings, one focused slash command, and any MCP server that connects Claude Code to your real source of truth. Add hooks and subagents after the workflow is stable.

Are MCP servers the same as Claude Code plugins?

MCP servers are one important plugin type. They connect Claude Code to external tools and data, while slash commands, hooks, subagents, settings, and output styles customize how Claude Code behaves inside a project.

How do I choose safe Claude Code plugins?

Prefer narrow permissions, project-scoped configuration, explicit verification steps, and read-only access until the workflow proves reliable. Avoid broad filesystem, token, or production access unless the task needs it.

Can these ideas work with Codex or Cursor too?

Many MCP server and workflow ideas can transfer to Codex CLI or Cursor, but Claude Code-specific slash commands, hooks, output styles, and subagents should be adapted to each agent's configuration model.

Related tools

Claude Code Settings Generator

Generate settings JSON with permission presets, environment variables, model defaults, and statusline config.

Open tool

Claude Code Subagent Generator

Create Claude Code subagent files with roles, triggers, model selection, and scoped tool permissions.

Open tool

MCP Server Config Generator

Build MCP client config for Claude Code, Cursor, Codex CLI, Claude Desktop, and VS Code.

Open tool