The 5 Best Claude Code Marketplace Alternatives in 2026

The Claude Code Marketplace is Anthropic's official mechanism for discovering and installing plugins, which are packages that can bundle skills, agents, hooks, MCP servers, slash commands, and LSP servers. Before plugins existed, teams shared scattered CLAUDE.md files, hook scripts, and MCP configs by hand; the marketplace turns those pieces into installable, shareable packages. The official marketplace (claude-plugins-official) is automatically available when you start Claude Code, so there's nothing to configure: run /plugin, browse the Discover tab, and install with a single command. Third-party marketplaces can be added with /plugin marketplace add owner/repo-name. Anthropic also maintains the open claude-plugins-official GitHub repository and a Knowledge Work Plugins collection aimed at knowledge workers in Claude Cowork. The strength is friction-free, first-party distribution inside the tool you already use. The tradeoff is that the marketplace is a distribution channel, not a vetting service: it doesn't publish per-plugin security scans, sandbox benchmarks, or quality scores, and it only serves Claude Code.

pricing See pricing

• + Zero setup: the official marketplace is available the moment you start Claude Code
• + First-party curation from Anthropic with a 32k-star official repo
• + Plugin packaging solves the old mess of scattered CLAUDE.md files, hooks and MCP configs
• + Refresh command (/plugin marketplace update) keeps the marketplace current

• − Claude Code only; no path to use the same skills in Codex, Cursor or Gemini CLI
• − No published security scans, sandbox benchmarks, or per-plugin quality scores
• − Installing many MCP-heavy plugins can choke the LLM context window, a caveat Composio documents

https://code.claude.com/docs/en/discover-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 multiplied faster than anyone's ability to judge them, and 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 and runs validation, license checks, security scans and quality judgments; 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 your agent only sees the skills and tools you toggled on; and Stay current tracks upstream changes, re-scans, re-benchmarks and notifies you. Each skill carries a 0-100 trust score built from security, quality, license, maintenance and community signals, with Lift always shown separately. Everything is free, funded by ads around blog articles that never influence scores.

pricing 0/month

• + Completely free, including the MCP connector, benchmarks and alerts
• + Addresses the ~26% vulnerability rate in agent skills with scans and SHA-pinned versions
• + Lift scores show whether a skill actually helps, tested in a sandbox with the same model
• + Works across Claude Code, Codex, Cursor, Gemini CLI and chat-app agents
• + Money never touches scores; ad revenue is kept separate from ratings and listings

• − Focused on skills rather than full plugin packages with hooks and commands
• − You need to configure the MCP connection and per-agent loadouts yourself, one step beyond a one-command plugin install

https://skillzbench.com

This is the open-source GitHub repository behind Anthropic's official marketplace, listed on the aitmpl plugin collections index with 32k stars. It's the Anthropic-managed directory of high-quality Claude Code plugins covering skills, agents, commands, hooks, and LSPs. Because it's the same collection the in-app Discover tab draws from, its practical role is as the browsable, cloneable mirror of the official marketplace: you can inspect plugin source before installing, add it explicitly as a marketplace with /plugin marketplace add, or keep it pinned for team use. Anthropic also publishes a companion Knowledge Work Plugins repository (23k stars) aimed at knowledge workers in Claude Cowork, which signals where the official ecosystem is heading beyond pure engineering. For teams that want first-party provenance and nothing else, this repo is the safest bet in the ecosystem. Its limits mirror the marketplace's: it's Claude Code-centric, and quality curation is not the same as published security scanning or benchmarking.

pricing See pricing

• + First-party provenance from Anthropic
• + Source code is fully inspectable on GitHub
• + Same collection powers the in-app Discover tab, so installs stay one command

• − Claude Code focused
• − Curation without published security scans or benchmark results
• − Smaller catalog than community mega-collections

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

Maintained by Composio, this curated list (875 stars on the aitmpl index) collects slash commands, subagents, MCP servers, and hooks for Claude Code, and it doubles as the on-ramp to Composio's integration platform. Its standout piece is Composio Connect MCP, installable straight from the repo, which lets Claude Code securely access apps like Jira, Linear, GitHub, and more through a single MCP server. Composio also solves a real problem the raw marketplace creates: installing several MCP servers (Linear, GitHub, Supabase, Vercel) loads every available tool into the context window and chokes it. Composio's dynamic tool loading brings tools in only as context requires, plus a remote workbench for chaining tools and a bash tool for handling large files outside the context window. You can start at dashboard.composio.dev in about a minute. The list itself is community-curated rather than benchmarked, so treat entries as a starting point, and the deeper value sits in Composio's platform.

pricing See pricing

• + Best-in-list approach to the context-window problem with dynamic tool loading
• + One MCP server replaces a pile of per-app MCP configs
• + Curated rather than a raw dump, with clear guidance on what to install first

• − Deeper features route through Composio's commercial platform
• − No benchmark or security-scan data on listed plugins
• − Primarily Claude Code oriented

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

With 22k stars, this is one of the largest cross-agent skill collections indexed by aitmpl: 345 Claude Code skills and agent skills, 30+ agents, 70+ custom commands, and 330+ skills with customizable references and scripts. Its defining trait is breadth of audience as much as breadth of count: it targets Claude Code, Codex, Gemini CLI, Cursor, and 8 more coding agents, and its categories span engineering, marketing, product, compliance, C-level advisory, research, business operations, commercial and finance, and daily productivity. For a small business, that means non-engineering staff can plausibly find skills relevant to their function, not just developers. The collection also includes hooks and commands alongside skills. The honest caveat is the same one that applies to every high-volume community collection: nothing here is benchmarked or security-scanned by the maintainer, so with research showing roughly 26% of agent skills carrying at least one vulnerability, a library this size deserves scrutiny before any skill reaches a production workflow.

pricing See pricing

• + Widest functional coverage in this comparison, well beyond engineering
• + Genuine cross-agent support across 12+ coding agents
• + Large, actively maintained collection with strong community traction

• − No vetting, scanning or benchmarking of individual skills
• − Volume makes discovery and quality judgment the user's job
• − Community-maintained, so consistency varies between skills

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

Buildwithclaude (3.2k stars) positions itself as 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. Where the mega-collections compete on volume, this one competes on orientation: it's a map of the ecosystem, useful when you know what capability you need (a hook, a skill, a plugin) but not which collection hosts the good version of it. Coverage spans the Claude family broadly, including Claude Desktop and the Agent SDK, which most plugin marketplaces ignore. For a small business evaluating the ecosystem for the first time, that makes it a reasonable first bookmark before committing to any single marketplace. The tradeoff is that a hub is only as good as what it links to: there's no original vetting layer, no benchmark data, and no security scanning, so anything you find here still needs its own quality check before it touches your workflows.

pricing See pricing

• + Clear, type-organized entry point into the whole Claude extension ecosystem
• + Broader Claude-family coverage than most marketplaces, including Desktop and Agent SDK
• + Lightweight: browse and jump straight to source repos

• − Aggregator, not a library: quality depends entirely on what it links to
• − No vetting, benchmarks or scans
• − Smaller community than the major collections

https://github.com/davepoon/buildwithclaude

This open-source marketplace (2.5k stars) is the volume play: 425 plugins, 2,810 skills, and 200 agents for Claude Code, distributed through the tonsofskills.com site and its own ccpi CLI package manager. The ccpi CLI is the differentiator: rather than browsing GitHub and running /plugin commands by hand, you get a package-manager workflow for discovering and installing from the catalog, closer to how developers already manage dependencies. For power users who want to stock a Claude Code setup quickly and don't mind sifting, the raw numbers are hard to ignore. The risks are the mirror image of the strengths: a catalog this size assembled at speed is exactly the environment where the ~26% vulnerability rate in agent skills bites hardest, and the project publishes no per-item security scans, quality scores, or benchmark results. If you use it, treat it as a candidate pool and verify before you trust, which is precisely the gap SkillzBench's free scanning and benchmarking fills.

pricing See pricing

• + Largest raw plugin and skill counts in this comparison
• + ccpi CLI gives a familiar package-manager experience
• + Open source and free to browse

• − No published security scanning, quality scores or benchmarks
• − Volume-first curation means wide quality variance
• − Claude Code only

https://github.com/jeremylongshore/Claude-Code-Plugins