The 5 Best Claude Code Marketplace Alternatives in 2026

Anthropic's official plugin marketplace for Claude Code. Plugins are packaged bundles that can include skills, agents, hooks, MCP servers, and other capabilities, replacing the old scatter of CLAUDE.md files, hook scripts, and custom slash commands. The marketplace is automatically available when you start Claude Code: you run /plugin, browse the Discover tab, and install with a single command such as /plugin install github@claude-plugins-official. Because Anthropic manages the official directory, it functions as the default, curated starting point for extending Claude Code, and refreshing it is one command (/plugin marketplace update claude-plugins-official). The catch is that it only exists inside the Claude Code ecosystem, and Claude Code itself has no free tier: the Pro plan runs $20/month, and DataCamp's 2026 comparison notes that sustained agentic use pushes developers toward the $100 or $200/month Max tiers. It also keeps you inside Anthropic's models.

pricing 20/month

• + Zero setup: Anthropic's docs say the official marketplace is automatically available when you start Claude Code
• + Anthropic-managed curation of high-quality plugins (32k stars on the official repo per aitmpl)
• + Plugin packaging replaces scattered CLAUDE.md files, hook scripts, and MCP configs
• + Deep ecosystem of first-party plugins including LSP integrations and knowledge-work plugins

• − Requires a paid Claude Code subscription; no free tier, and heavy agent use pushes toward $100-$200/month
• − Locks you into Anthropic's models and a terminal-first workflow
• − No published per-skill security scanning or benchmark results, notable given research showing ~26% of agent skills carry at least one vulnerability

https://github.com/anthropics/claude-plugins-official

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. Its premise: SKILL.md instruction packages for Claude Code, Codex, Cursor, Gemini CLI, and chat-app agents are everywhere, but nobody could tell which ones actually help and which are risky, and published research found about 26% of skills carry at least one vulnerability. SkillzBench runs five pillars today: Discover (continuous crawling and deduping of skills from GitHub), Verify (pinning every version to a commit SHA, validating, license-checking, security-scanning, and quality-judging), Benchmark (testing every skill in a sandbox with the same model, with and without the skill, and publishing the Lift in percentage points), Equip (one free MCP server plus per-agent loadouts, so the agent only sees what you toggled on), and Stay current (tracking upstream changes, re-scanning, re-benchmarking, and notifying). A 0-100 trust score combines security, quality, license, maintenance, and community signals, and Lift is always shown separately from ratings. Advertising in blog articles funds the platform; money never touches scores.

pricing 0

• + Completely free, including the MCP connector and benchmarks; ad-funded, and ad money never influences scores
• + Addresses the ~26% vulnerability rate in agent skills with commit-SHA pinning and security scanning
• + Lift scores show measured performance impact, separate from ratings, so you can see what a skill actually does
• + Works across agents (Claude Code, Codex, Cursor, Gemini CLI), not just one ecosystem

• − It is a directory and connector, not an in-editor plugin installer, so day-to-day installs in Claude Code still happen through plugin commands
• − Benchmark methodology is single-model by design, so Lift reflects one model context rather than every model you might use
• − Younger and less well-known than the official Anthropic marketplace

https://skillzbench.com

Awesome Claude Plugins is Composio's community marketplace and one of the most-cited third-party sources for Claude Code extensions. It is a curated GitHub repository covering slash commands, subagents, MCP servers, and hooks, and it installs directly into Claude Code with a single command: /plugin marketplace add owner/repo-name. The collection is tightly integrated with Composio's broader tooling story; the repo includes Composio Connect MCP, which lets Claude Code securely access apps like Jira, Linear, and GitHub from a single MCP server, addressing the context-window problem that Composio identifies when you stack many MCP servers at once (each loads all its available tools into the context). For developers who want to go beyond Anthropic's official catalog without leaving the /plugin workflow, it is the natural first stop, and Composio's 2026 plugin guide uses it as the reference example for third-party marketplace installs.

pricing See pricing

• + One-command install inside the native Claude Code plugin flow
• + Backed by Composio, which actively documents and maintains the ecosystem
• + Solves the multi-MCP context-window choke with dynamic tool loading

• − Community curation rather than vendor vetting; no published per-plugin security scans
• − Tied to Composio's ecosystem, which adds a vendor relationship beyond Anthropic

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

aitmpl.com maintains a browsable catalog of the most popular Claude Code plugin collections, skill registries, and component marketplaces on GitHub, indexing 34 collections as of 2026. Each entry shows the maintainer, star count, and capability tags (skills, agents, commands, hooks, LSPs, MCPs), which makes it the fastest way to compare the shape of the ecosystem at a glance: ECC by affaan-m leads with 229k stars, Claude Mem by thedotmack sits at 87k, Anthropic's official directory at 32k, and Claude Hud at 26k. Specialized entries cover everything from Playwright browser-automation skills to HashiCorp's official agent skills collection and a Timescale Postgres guide plugin. It is a discovery layer rather than an installer: you evaluate options on the site, then add the underlying GitHub repo as a marketplace in Claude Code. For small teams surveying what exists before committing, that bird's-eye view is the value.

pricing See pricing

• + Broadest single view of the marketplace ecosystem, including star counts for popularity signals
• + Surfaces specialized collections (Playwright automation, Postgres guidance, equity research workflows)
• + Tag system makes it easy to filter by capability type

• − Discovery only; you still install through GitHub marketplace commands
• − No quality scoring or security verification of listed collections

https://www.aitmpl.com/plugins/

Buildwithclaude by davepoon (3.2k stars per aitmpl) is a single hub for finding Claude Skills, Agents, Commands, Hooks, Plugins, and Marketplace collections. Its differentiator is scope: it covers Claude Code, Claude Desktop, the Agent SDK, and OpenClaw, so it is useful if your team's agent usage stretches beyond the terminal CLI. Where the official marketplace optimizes for a curated short list, Buildwithclaude optimizes for being the map of everything, organized by artifact type so you can jump straight to the hooks or skills section you need. The tradeoff is the same one every hub-style directory carries: listings are indexed rather than vetted, so quality varies by source repo, and there is no published security or benchmark layer. It fits developers who already know what capability they need and want to find which collection provides it across several Claude surfaces.

pricing See pricing

• + Widest Claude-surface coverage in one place, including Desktop and the Agent SDK
• + Organized by artifact type, which speeds up targeted searches
• + Free and open on GitHub

• − Hub-style listing with no quality scoring
• − 3.2k stars means a smaller community than the biggest collections

https://github.com/davepoon/buildwithclaude

Claude Code Plugins Plus Skills by jeremylongshore is the volume play in the ecosystem: 425 plugins, 2,810 skills, and 200 agents for Claude Code, distributed through an open-source marketplace at tonsofskills.com with its own CLI package manager called ccpi (2.5k stars per aitmpl). Where curated marketplaces hand-pick a short list, this one maximizes catalog breadth and adds a package-manager workflow, so teams can script installs the way they would with npm or pip. That scale cuts both ways. A catalog of nearly three thousand skills inevitably includes entries of wildly varying quality, and the research shows no per-skill security scanning or benchmarking layer, which matters given published findings that roughly 26% of agent skills carry at least one vulnerability. It suits power users who want maximum choice and are comfortable vetting entries themselves.

pricing See pricing

• + Largest catalog in this comparison by raw numbers
• + ccpi CLI brings package-manager ergonomics to skill installs
• + Open source and self-hostable workflow

• − Volume-first approach means uneven quality across nearly 3,000 skills
• − No published security scanning or benchmarking of individual entries

https://tonsofskills.com

Superpowers Marketplace by obra (1.1k stars per aitmpl) is a deliberately curated Claude Code plugin marketplace covering commands, MCP servers, and skills. Its philosophy is the opposite of the volume catalogs: a smaller, hand-picked set where the curation itself is the product. It installs as a standard third-party marketplace inside Claude Code's /plugin flow, so adopting it costs one command. The smaller catalog means you may not find niche capabilities like the Playwright browser-automation skill or specialized LSP integrations that bigger indexes carry, but what is listed tends to be coherent and maintained by a community that cares about workflow quality. For a small business team that wants a short, sane list of extensions rather than a firehose, that restraint is a feature. Like the other community marketplaces, it publishes no per-skill security scans or benchmark results.

pricing See pricing

• + Curation-first, so the listed set is coherent and maintained
• + Simple adoption through the native plugin command
• + Free and open source

• − Small catalog (1.1k stars) may lack niche capabilities
• − No formal security or benchmark verification layer

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