The 5 Best Claude Code Marketplace Alternatives in 2026

Anthropic's official marketplace for Claude Code plugins. According to Anthropic's docs, plugins are bundles that can include skills, agents, hooks, MCP servers and other capabilities, and the marketplace lets you discover and install them without building them yourself. It's available automatically when you start Claude Code: run /plugin, browse the Discover tab, and install directly from the official claude-plugins-official repo. Before plugins, teams shared scattered CLAUDE.md files, hook scripts, MCP configs and slash commands; the marketplace packages those pieces into installable units. Popular plugin categories include LSP code-intelligence plugins (TypeScript, Pyright, Rust Analyzer, Go), MCP integrations for GitHub, Linear, Slack, Vercel, Sentry and Supabase, code review plugins like CodeRabbit and PR Review Toolkit, and security plugins such as Security Guidance and Semgrep. The trade-off: it serves Claude Code only, and curation is Anthropic's, with no published per-skill security scans or measured performance Lift.

pricing See pricing

• + Zero setup: the official marketplace is available the moment you start Claude Code
• + Anthropic-managed, so the core directory is curated by the vendor itself
• + Plugin packaging replaces scattered CLAUDE.md files, hooks and MCP configs
• + Rich ecosystem of LSP, MCP, code review and security plugins documented by Composio

• − Claude Code only; no coverage for Codex, Cursor or Gemini CLI
• − No published per-skill security scans, version pinning or benchmark Lift scores
• − Using Claude Code itself requires a subscription; the Pro plan is $20/month and heavy agent use pushes toward $100-$200/month Max tiers (DataCamp)
• − Loading many MCP servers at once can crowd the LLM context window

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. 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 has not: published research found roughly 26% of skills carry at least one vulnerability. SkillzBench runs five pillars today. 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 a quality judgment. Benchmark tests every skill in a sandbox with the same model, with and without the skill, and publishes the Lift in percentage points. Equip gives you one free MCP server plus per-agent loadouts, so you toggle skills and tools 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, and Lift is always shown separately from ratings; money never touches scores. Everything is free, funded by ads in blog articles.

pricing 0

• + Everything is free, and ad revenue never influences scores or listings
• + Only option here that benchmarks every skill and publishes the measured Lift
• + Security scanning and SHA version pinning address the ~26% vulnerability rate found in published research
• + Works across Claude Code, Codex, Cursor, Gemini CLI and chat-app agents through one MCP server
• + Per-agent loadouts prevent the context-window bloat of loading every MCP server at once

• − Not a native Claude Code /plugin marketplace, so installation flows differ from Anthropic's built-in Discover tab
• − Focused on skills rather than full plugin bundles with hooks and LSP servers
• − Newer and smaller community than the official Anthropic directory

https://skillzbench.com

Awesome Claude Plugins is Composio's open-source collection of Claude Code plugins, installable as a third-party marketplace with /plugin marketplace add or by cloning the GitHub repo. Its standout contribution is Composio Connect MCP, which lets Claude Code securely access apps like Jira, Linear, GitHub and more from a single MCP server. Composio's write-up also highlights a real operational problem: installing many MCP servers (Linear, GitHub, Supabase, Vercel) loads all their tools into the LLM context window and chokes it. Composio addresses that with dynamic tool loading (tools load as needed), a remote workbench for chaining tools into workflows, and a bash tool for handling large files outside the context window. Setup is advertised as a minute or less via dashboard.composio.dev. It's a strong pick when your Claude Code work depends on external services, though it's a curated list rather than a vetted registry, and it doesn't publish per-skill benchmarks.

pricing See pricing

• + Solves the multi-MCP context-window problem with dynamic tool loading
• + Strong app-integration focus for issue trackers, repos and deploys
• + Free and open source

• − Requires manual third-party marketplace setup rather than native discovery
• − Curated list without per-skill security scans or benchmarks
• − Claude Code-centric

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

aitmpl.com maintains a browsable directory of Claude Code plugin collections, skill registries and component marketplaces on GitHub, currently listing 34 collections. Each entry shows the maintainer, approximate GitHub stars and capability tags (skills, agents, commands, hooks, MCPs, LSPs), so you can size up a collection's popularity and scope before installing it. The catalog spans everything from Anthropic's official repo (about 32k stars) and huge community collections like aecc (229k stars) and Claude Mem (87k stars) to niche entries like HashiCorp's agent skills collection and the Timescale Postgres guide plugin. It's a meta-directory: it doesn't host plugins itself, it points you to the underlying GitHub repos where you install them. That makes it a fast way to survey the ecosystem in one tab, but it offers no vetting layer; star counts are popularity signals, not quality guarantees, and there's no security scanning or benchmark data attached to listings.

pricing See pricing

• + Broadest single-page view of the Claude Code plugin ecosystem
• + Free and requires no install to browse
• + Covers collections that also target Codex, OpenCode and Cursor

• − Meta-directory only: no hosting, no vetting, no benchmarks
• − Star counts can mislead on quality
• − You still handle installation repo by repo

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

Claude Skills by alirezarezvani is one of the largest single-community collections in the ecosystem, with roughly 22k GitHub stars per the aitmpl listing. It packages 345 Claude Code skills and agent skills, 30+ agents and 70+ custom commands, plus customizable references and scripts, aimed at Claude Code, Codex, Gemini CLI, Cursor and 8 more coding agents. Where most collections lean engineering-heavy, this one deliberately spans business functions: engineering, marketing, product, compliance, C-level advisory, research, business operations, commercial and finance, and daily productivity. For a small business adopting AI agents beyond pure coding, that breadth is the appeal: one install point for skills that draft marketing copy, run research, or handle operational checklists. The trade-offs are the usual for community collections: quality varies skill by skill, there's no published security scanning or benchmark Lift, and a single maintainer's curation standards are all the assurance you get.

pricing See pricing

• + Very broad skill library including non-engineering business functions
• + Works with many coding agents, not just Claude Code
• + Free and open source

• − No per-skill security scans, license checks or benchmark Lift
• − Single-maintainer curation
• − Quality varies across 345 skills

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

Claude Code Plugins Plus Skills by jeremylongshore is a volume-first open-source marketplace listing 425 plugins, 2,810 skills and 200 agents for Claude Code, distributed through tonsofskills.com with its own ccpi CLI package manager. The ccpi CLI is the differentiator: instead of adding a marketplace repo and browsing in Claude Code, you install packages from the terminal, which suits developers who script their setups or manage several machines. The collection's scale makes it a useful bulk resource when you want to experiment widely, and the dedicated website gives it more of a product surface than a plain GitHub list. The caveats mirror its strengths: at 2,810 skills, per-skill quality assurance is impractical, there's no published security scanning or benchmarking layer, and coverage is Claude Code-focused. Treat it as a large raw inventory rather than a vetted catalog.

pricing See pricing

• + Largest raw inventory in this comparison
• + ccpi CLI enables scripted, repeatable installs
• + Free and open source

• − No verification, scanning or benchmark layer across thousands of skills
• − Claude Code-centric
• − Volume over curation

https://tonsofskills.com

Buildwithclaude by davepoon (about 3.2k GitHub stars per the aitmpl listing) is a single hub for finding Claude skills, agents, commands, hooks, plugins and marketplace collections that extend Claude Code, Claude Desktop, the Agent SDK and OpenClaw. Its value is consolidation: rather than hopping between scattered GitHub repos and blog posts, you get one indexed starting point covering the full range of extension types, including collections that themselves aggregate other marketplaces. For someone setting up Claude Code for the first time, it works as a guided tour of what's possible before you commit to any particular collection. It's a discovery hub, though, not a registry: it doesn't host the extensions, doesn't scan them for security issues, doesn't pin versions and doesn't publish performance data. Once you find something interesting, verification is on you.

pricing See pricing

• + Good first-stop orientation for the Claude extension ecosystem
• + Covers Claude Desktop and the Agent SDK, which most lists skip
• + Free

• − Discovery only: no vetting, scanning or benchmarks
• − Smaller catalog than the big collections
• − No cross-agent MCP tooling

https://github.com/davepoon/buildwithclaude