The 5 Best Claude Code Marketplace Alternatives in 2026

The official plugin and skill directory for Claude Code, managed by Anthropic. Plugins are packages that can bundle skills, agents, hooks, MCP servers, and slash commands, replacing the old scatter of CLAUDE.md files, hook scripts, and custom configs teams used to share by hand. The marketplace is available automatically when you start Claude Code: run /plugin, browse the Discover tab, and install with a single command such as /plugin install github@claude-plugins-official. Anthropic also publishes the claude-plugins-official repository (32k GitHub stars) and a Knowledge Work Plugins collection aimed at non-developers using Claude Cowork. If a plugin goes missing, /plugin marketplace update claude-plugins-official refreshes the directory. You can add third-party marketplaces the same way, which turns the marketplace into a framework for community collections rather than a closed store. The trade-off: vetting is Anthropic's own, there are no published per-skill security scans or benchmarks, and nothing here works outside Claude Code.

pricing See pricing

• + Zero setup: the official marketplace is available the moment you start Claude Code
• + One-command installs and a refresh command to keep the directory current
• + Anthropic-curated, so the official listings meet a consistent quality bar
• + Open to third-party marketplaces, so community collections plug into the same workflow

• − Works only inside Claude Code; teams also running Codex, Cursor, or Gemini CLI need separate sources
• − No published per-skill security scans, benchmarks, or independent trust scores
• − Practical use requires a Claude Code subscription, and heavy agent use pushes toward $100-$200/month Max plans
• − Smaller catalog than the largest community skill collections

https://code.claude.com/docs/en/plugins

SkillzBench is AIQ Labs' free, independent directory of AI agent skills plus a free MCP tool that connects any agent to the skills you choose. It exists because agent skills (SKILL.md instruction packages for Claude Code, Codex, Cursor, Gemini CLI, and chat-app agents) have exploded in number while quality control hasn't: published research found roughly 26% of skills carry at least one vulnerability. SkillzBench answers that with five pillars. Discover continuously crawls and dedupes skills from GitHub. Verify pins every version to a commit SHA, validates it, checks the license, runs a security scan, and judges quality. Benchmark tests every skill in a sandbox with the same model, with and without the skill, and publishes the Lift in percentage points. Equip provides one free MCP server plus per-agent loadouts, so you toggle skills per agent and the agent only sees what you chose. Stay current tracks upstream changes, re-scans, re-benchmarks, and notifies you. A 0-100 trust score combines security, quality, license, maintenance, and community signals; Lift is always shown separately, and money never touches scores. Everything is free, funded by blog advertising.

pricing See pricing

• + Everything is free: directory, MCP connector, benchmarks, and alerts, funded by blog ads rather than users
• + Only option here that independently security-scans and benchmarks each skill and publishes the measured Lift
• + Works across Claude Code, Codex, Cursor, and Gemini CLI through one MCP connector
• + Per-agent loadouts limit each agent to exactly the skills you chose
• + Version pinning to commit SHAs protects you from silent upstream changes

• − It's a separate website and MCP setup rather than a command inside your editor
• − Ad-funded model means ads appear around blog content, though never in scores or listings
• − You need to check skillzbench.com for current catalog size rather than relying on a fixed published count

https://skillzbench.com

Awesome Claude Plugins is the community-curated list from Composio covering slash commands, subagents, MCP servers, and hooks for Claude Code. It doubles as an installable third-party marketplace: you add it to Claude Code with /plugin marketplace add owner/repo-name, or clone the GitHub repo and install directly from the source directory. The Composio write-up behind it also packages a Connect MCP that lets Claude Code securely reach apps like Jira, Linear, and GitHub from a single MCP server, which addresses the context-window problem that hits you when you stack many MCP servers at once. As an awesome-list style resource, its strength is breadth and active community maintenance; its weakness is that curation means 'selected and described,' not 'tested.' There are no published security scans or per-skill benchmarks, so you're relying on the maintainers' judgment and community visibility.

pricing See pricing

• + Free and open source, cloneable straight from GitHub
• + Broad coverage across commands, subagents, MCPs, and hooks
• + Backed by Composio's dynamic tool loading, which loads tools only as context requires
• + Installs through the standard marketplace workflow you already know

• − No independent security scanning or benchmarking of listed plugins
• − Claude Code focused, so it doesn't help multi-agent teams
• − Quality varies by contributor, as with any community list

https://github.com/composiohq/awesome-claude-plugins

This 22k-star GitHub collection is the volume play: 345 Claude Code skills and agent skills, 30+ agents, 70+ custom commands, plus customizable references and scripts. Where it genuinely differs from most lists is reach: it targets Claude Code, Codex, Gemini CLI, Cursor, and 8 more coding agents, and its skills span engineering, marketing, product, compliance, C-level advisory, research, business operations, commercial and finance, and daily productivity. For a small business, that breadth matters, because the marketing or finance skills are the ones a non-developer owner might actually use. The trade-off is the flip side of volume: one maintainer's collection with no published security scans, license checks, or benchmark results per skill. With roughly a quarter of agent skills carrying at least one vulnerability per published research, installing from a large unverified pack means doing your own vetting, at least for anything that touches sensitive data or runs commands.

pricing See pricing

• + Largest single-collection breadth of the options compared here
• + Covers business functions small businesses actually need, not just coding
• + Works across many coding agents, not only Claude Code
• + Free and open source with a large star count signaling active interest

• − No published security scans, benchmarks, or trust scores per skill
• − Single-maintainer collection; maintenance depends on one person
• − Sheer size makes finding the right skill slower

https://github.com/alirezarezvani/claude-skills

Buildwithclaude (3.2k GitHub stars) is a single hub for finding Claude skills, agents, commands, hooks, plugins, and marketplace collections, and it extends beyond the CLI to Claude Desktop, the Agent SDK, and OpenClaw. Think of it as a discovery layer rather than a marketplace itself: it organizes the ecosystem so you can find what exists, then install through Claude Code's normal mechanisms. That positioning has real value for someone new to the plugin ecosystem who doesn't yet know whether they need a skill, a hook, or an MCP server; the hub explains the categories and links to the collections. The limitation follows from the model: aggregation isn't verification. Buildwithclaude doesn't security-scan, benchmark, or version-pin the things it links to, and its Claude-centric scope means Codex or Gemini CLI users get little from it. It's best treated as a map of the territory, paired with a verification source for anything you actually install.

pricing See pricing

• + Good orientation for newcomers who don't know which extension type they need
• + Broader Claude-ecosystem coverage than CLI-only lists
• + Free and open source

• − Aggregates links; no scanning, benchmarking, or pinning of underlying skills
• − Claude ecosystem only
• − Smaller community (3.2k stars) than the biggest collections

https://github.com/davepoon/buildwithclaude

Superpowers Marketplace (1.1k stars, by obra) is a curated Claude Code plugin marketplace covering commands, MCPs, and skills. It sits in the same pattern as the other third-party marketplaces: add it with /plugin marketplace add, browse its curated set, install what fits. Its appeal is a focused, opinionated selection from a maintainer with a visible track record in the Claude community, which some users prefer over sprawling 300-skill packs. The same caveats apply as to any curated list: curation reflects the maintainer's judgment, not independent testing, and there are no published security scans or benchmark results. Its catalog is also the smallest of the options compared here, and it's Claude Code only. If you like the maintainer's taste, that focus is a feature; if you need breadth or evidence, look elsewhere.

pricing See pricing

• + Focused, opinionated curation from a known maintainer
• + Free and installs through the normal /plugin workflow
• + Small enough to actually review everything on offer

• − Smallest catalog in this comparison (1.1k stars)
• − No independent security or benchmark verification
• − Claude Code only

https://github.com/obra/superpowers-marketplace