19.7K listings indexed
Every MCP server, skill, and plugin — with the receipts.
Claude SuperPower indexes the Claude Code and MCP ecosystem and shows you what other directories don't: what each tool can actually do, what it costs your context, how well maintained it is, and whether anyone has reviewed it.
- Listings
- 19.7K
- Trust-scored
- 1.3K
- Component pages
- 3.2K
- Reviews
- 1
Browse the ecosystem
Best of, by category →Plugins
308Bundles that ship skills, MCP servers, hooks, and subagents together as one install.
Browse pluginsMCP Servers
16.2KModel Context Protocol servers that connect Claude to tools, APIs, and data.
Browse mcp serversSkills
2.6KPackaged, reusable capabilities Claude can invoke to complete specific tasks.
Browse skillsSubagents
230Specialized agent configurations for delegated, focused work.
Browse subagentsHooks
149Event-driven scripts that run automatically at points in the Claude Code lifecycle.
Browse hooksCommands
249Slash commands that trigger predefined workflows in Claude Code.
Browse commandsStart from what you're doing
All use cases →Categories tell you what something is. These tell you what it's for — and each one spans MCP servers, plugins, skills, and commands together.
What you get here and nowhere else
Most directories stop at a name, a description, and a star count. These are the things we compute from each project's own source.
Trust you can audit
Six named factors, each with its points and the reason it scored that way. Not a black-box number.
Real permission manifests
Exactly what a plugin can do — "run shell (7 components), write files (3)" — read from its own frontmatter.
Context cost up front
How many tokens a plugin consumes just by being enabled, before you invoke anything.
Side-by-side comparison
Compare any two listings on stars, rating, licence, and maintenance in one view.
Community reviews
Ratings from people who actually ran it — the one thing no other directory in this space offers.
Component-level search
Search skills, subagents, and commands directly instead of hunting through plugin bundles.
Most starred
See all →
superpowers
Superpowers teaches Claude brainstorming, subagent driven development with built in code review, systematic debugging, and red/green TDD. Additionally, it teaches Claude how to author and test new skills.
ECC
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
ECC
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
Anthropic Example Skills
Public repository for Agent Skills
gstack
Use Garry Tan's exact Claude Code setup: 23 opinionated tools that serve as CEO, Designer, Eng Manager, Release Manager, Doc Engineer, and QA
ui-ux-pro-max-skill
An AI SKILL that provide design intelligence for building professional UI/UX multiple platforms
ponytail
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
Official MCP Servers Collection
Model Context Protocol Servers
Actively maintained
by upstream commit date
agentops
The operating loop a coding agent follows — and skills to orchestrate multi-agent systems.
oh-my-claudecode
Teams-first Multi-agent orchestration for Claude Code
open-design
🎨 The open-source Claude Design alternative. 🖥️ Local-first desktop app. 🖼️ Your coding agent becomes the design engine: prototypes, landing pages, dashboards, slides, images & video — real files, HTML/PDF/PPTX/MP4 export. 🤖 Claude Code / Codex / Cursor / Gemini / OpenCode / Qwen & 20+ CLIs via BYOK.
raptor
Raptor turns Claude Code into a general-purpose AI offensive/defensive security agent. By using Claude.md and creating rules, sub-agents, and skills, and orchestrating security tool usage, we configure the agent for adversarial thinking, and perform research or attack/defense operations.
Frequently asked questions
What is an MCP server?+
An MCP (Model Context Protocol) server is a small program that exposes tools, data, or APIs to an AI coding agent through a standard protocol. Connecting one lets the agent read from and act on that system directly — querying a database, searching the web, or opening pull requests. MCP is an open standard, so the same server works in Claude Code, Cursor, Windsurf, and other hosts.
What is a Claude skill?+
A skill is a packaged, reusable capability the agent loads on demand to perform a specific task well — a folder of instructions plus optional scripts. Skills keep specialised know-how out of your prompt until it is actually needed, which is why their context cost matters.
What is a Claude Code plugin?+
A plugin is a bundle. One install can ship skills, MCP servers, hooks, subagents, and slash commands together, so you get a whole workflow in a single step instead of wiring up each piece yourself. Every plugin page here lists exactly what is inside it.
How do you decide what is trustworthy?+
Every listing gets a trust score built from six named factors — auditable source, documentation, maintenance, adoption, licence, and provenance — and the full arithmetic is shown on the page so you can check it. It measures auditability and maintenance, not code safety: it is not a security audit.
Is Claude SuperPower affiliated with Anthropic?+
No. Claude SuperPower is an independent directory. Listings are sourced from public GitHub repositories and the official MCP registry, and anything unauditable or self-declared as non-functional is held for human review rather than published.
Get the weekly digest
New MCP servers, skills, and plugins worth knowing about. One email a week, no spam.
Built something?
Paste a GitHub repo and we'll pull the details, score it, and list it. An editor reviews every submission before it goes live.
Submit a listing