The 5 Best Claude Code Marketplace Alternatives in 2026

Anthropic's official directory for Claude Code plugins. A plugin is a package that can bundle skills, agents, hooks, MCP servers, slash commands, and LSP servers, and the marketplace is where you discover and install them without building each piece yourself. It 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. The official marketplace repo (anthropics/claude-plugins-official) is Anthropic-managed, and third-party marketplaces can be added with /plugin marketplace add owner/repo-name. For teams already standardized on Claude Code, it is the lowest-friction way to package shared workflows that used to live in scattered CLAUDE.md files, hook scripts, and MCP configs. The trade-off is that it is a Claude Code-only channel: nothing here helps a business running Codex, Cursor, or Gemini CLI, and the directory does not publish independent security scans or performance benchmarks for individual plugins.

pricing See pricing

• + Zero setup: the official marketplace is available the moment you start Claude Code
• + Officially curated by Anthropic, so the baseline quality bar is higher than random GitHub dumps
• + One-command installs and a marketplace refresh command keep plugins current
• + Free to browse and install

• − Claude Code only; no value for teams on Codex, Cursor, or Gemini CLI
• − No published security scans or benchmark results per plugin
• − Claude Code itself has no free tier, and heavy agent use pushes you toward $100-$200/month Max plans
• − Anthropic-only model lock-in comes with the territory

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. Agent skills (SKILL.md instruction packages for Claude Code, Codex, Cursor, Gemini CLI, and chat-app agents) are everywhere, but published research found roughly 26% of them carry at least one vulnerability, and no marketplace tells you which ones actually help. SkillzBench runs five pillars: Discover continuously crawls and dedupes skills from GitHub; Verify pins every version to a commit SHA and validates, license-checks, security-scans, and quality-judges each one; 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 the agent only sees the skills and tools you toggled on; and Stay current tracks upstream changes, re-scans, re-benchmarks, and notifies you. Trust scores run 0-100 from security, quality, license, maintenance, and community signals, and Lift is always shown separately from ratings. Everything is free, funded by advertising around blog articles that never influences scores or listings.

pricing 0

• + Completely free, including the MCP connector, benchmarks, and alerts
• + Addresses the safety gap directly: skills are security-scanned and version-pinned before you touch them
• + Benchmark Lift gives you evidence a skill actually helps, not just that it exists
• + Works across Claude Code, Codex, Cursor, Gemini CLI, and chat-app agents, not one vendor's tool
• + Per-agent loadouts prevent skills and tools from leaking into agents that shouldn't see them

• − An independent directory, so it lacks the official-vendor cachet of Anthropic's own marketplace
• − Not a plugin package manager; you equip skills through its MCP rather than /plugin install

https://skillzbench.com

Awesome Claude Plugins is Composio's open-source, community-curated marketplace of Claude Code plugins, installable directly into Claude Code with /plugin marketplace add composiohq/awesome-claude-plugins. The repo also ships Composio Connect MCP, which lets Claude Code securely access apps like Jira, Linear, GitHub, and more from a single MCP connection. Composio's write-up highlights a real pain point with stacking MCP servers: installing Linear, GitHub, Supabase, and Vercel MCPs together loads every available tool into the context window and chokes it. Composio addresses that with dynamic tool loading (tools load as needed based on context), a remote workbench for chaining tools into workflows, and a bash tool for handling large files outside the context window. As a marketplace, it is free and genuinely useful for Claude Code users, but curation is a human list, not a security pipeline: there are no published per-plugin scans, license checks, or benchmark results.

pricing See pricing

• + Free and open source
• + Curated by an active team rather than an unsorted dump
• + Pairs with Composio's dynamic tool loading, which solves the multi-MCP context-window problem

• − Claude Code focused
• − No per-plugin security scanning or benchmark data
• − Getting started with Composio's broader platform means another dashboard (dashboard.composio.dev)

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

Buildwithclaude (by davepoon, around 3.2k stars) 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: instead of hunting across scattered GitHub repos for each extension type, you get one categorized index covering the full range of what can be installed. That breadth is also its limitation. It is an aggregator, not a verifier. Listings are organized by type, but there is no per-item security scan, license check, version pinning, or performance benchmark, so you are trusting the underlying repo authors directly. For a small business exploring what the Claude ecosystem offers before committing, it is a good map. For deciding whether a specific skill is safe to run against your codebase, you will need to do that diligence yourself or use a service that does it for you.

pricing See pricing

• + Free and open source
• + Broadest type coverage of any Claude-focused hub in this list
• + Includes OpenClaw and Agent SDK, not just the CLI

• − Aggregates rather than verifies; no scans, benchmarks, or quality scores
• − Claude ecosystem only
• − Quality varies by underlying repo author

https://github.com/davepoon/buildwithclaude

Claude Code Plugins Plus Skills (by jeremylongshore, around 2.5k stars) is an open-source marketplace listing 425 plugins, 2,810 skills, and 200 agents for Claude Code, with its own ccpi CLI package manager and a companion site at tonsofskills.com. Raw volume is the pitch: no other Claude-focused collection in this comparison comes close on catalog size. Volume cuts both ways. A catalog that large cannot be hand-curated in any meaningful sense, and the listing publishes counts rather than quality signals: no per-skill security scanning, no license verification, no benchmark results, no trust scoring. If you know exactly what you need and want to grep a big index, it works. If you are a small business owner trying to figure out which of 2,810 skills are safe and effective, the catalog size becomes the problem rather than the feature.

pricing See pricing

• + Free and open source
• + Largest raw catalog in this comparison
• + CLI installer makes bulk browsing and installing faster

• − No published security, license, or quality vetting per item
• − Claude Code only
• − Catalog size makes discovery without quality signals difficult

https://tonsofskills.com

Superpowers Marketplace (by obra, around 1.1k stars) is a curated Claude Code plugin marketplace covering commands, MCP servers, and skills. It sits in the middle ground of this list: smaller and more opinionated than the mega-catalogs, more community-driven than Anthropic's official directory. Curation is the product. Someone has looked at what's included, which matters more than it sounds when the alternative is an unsorted 2,810-item dump. But curation is a judgment call, not a verification pipeline. There are no published security scans, version pinning, license checks, or benchmark results, and the collection is Claude Code focused. If you like the maintainer's taste and work in Claude Code, it's a pleasant place to shop. If you need evidence, it doesn't provide any.

pricing See pricing

• + Free and open source
• + Human-curated, so the signal-to-noise is better than raw dumps
• + Covers commands, MCPs, and skills in one marketplace

• − Small catalog compared with the volume players
• − No security scanning or benchmark data
• − Claude Code only

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

Ecc (by affaan-m, around 229k stars, the most popular collection in the aitmpl index) describes itself as an agent harness performance optimization system: skills, instincts, memory, security, and research-first development for Claude Code, Codex, OpenCode, Cursor, and beyond. Unlike the directory-style entries here, Ecc is a single opinionated system you adopt wholesale. Its multi-agent support is a genuine differentiator, and its popularity suggests the approach resonates with a large developer audience. The trade-offs are the flip side of being opinionated: you take the whole harness, not just the pieces you want, and despite shipping security features, it publishes no independent per-skill verification or benchmark results. For a small business, adopting a 229k-star harness built for individual power developers is a bigger commitment than picking verified skills a la carte.

pricing See pricing

• + Free and open source
• + Most popular collection in its category (229k stars)
• + Genuine multi-agent support across four or more coding agents

• − All-or-nothing harness rather than a pick-and-choose directory
• − No independent per-skill verification or benchmarks
• − Built for individual developers, not small-business team workflows

https://github.com/affaan-m/ecc

Claude Skills by alirezarezvani (around 22k stars) is a large multi-domain collection: 345 Claude Code skills and agent skills spanning 30+ agents, 70+ custom commands, and 330+ skills, with customizable references and scripts. It targets Claude Code, Codex, Gemini CLI, Cursor, and eight more coding agents, and its domain coverage is unusually broad for a skills repo: engineering, marketing, product, compliance, C-level advisory, research, business operations, commercial and finance, and daily productivity. For a small business, that business-function coverage is the interesting part; most competing collections are developer-only. The caveats mirror the other community catalogs: counts are published, but there is no per-skill security scanning, license checking, version pinning, or benchmarking, and with hundreds of skills across dozens of domains, quality inevitably varies by author.

pricing See pricing

• + Free and open source
• + Broad business-domain coverage (marketing, compliance, finance, operations), rare among skills repos
• + Multi-agent compatibility across 12+ coding agents

• − No published security, license, or quality vetting per skill
• − Quality varies across hundreds of contributed skills
• − No benchmark data to show which skills actually help

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