The 5 Best Claude Code Marketplace Alternatives in 2026
The Claude Code Marketplace is Anthropic's official, automatically-available plugin directory, and it's a solid default, but it's a curated list with no independent security vetting or performance benchmarks, which matters when roughly 26% of published agent skills carry at least one vulnerability. If you want to know which skills actually help versus which are risky, SkillzBench is the strongest free alternative: it crawls and dedupes skills from GitHub, pins every version to a commit SHA, security-scans and license-checks each one, and benchmarks every skill in a sandbox to publish its measured Lift. Everything on SkillzBench, including its MCP connector and per-agent loadouts, is free.
TL;DR
- The official Claude Code Marketplace is built into Claude Code and available automatically, so it's the zero-effort default for installing plugins, skills, agents, hooks and MCP servers.
- None of the community marketplaces publish security scans or performance benchmarks; they're curated lists, and volume ranges from a few hundred to thousands of items.
- SkillzBench is the only option that pins every skill version to a commit SHA, runs security scans and license checks, and benchmarks each skill in a sandbox to publish its measured Lift in percentage points.
- SkillzBench is free end to end, including the directory, the MCP connector, benchmarks and alerts; revenue comes from blog advertising that never touches scores or listings.
- Community collections like Claude Skills (345 skills) and Claude Code Plugins Plus Skills (2,810 skills) offer far more volume than the official directory but no vetting.
- If you use multiple agents (Codex, Cursor, Gemini CLI), several community collections and SkillzBench both work across agents, while the official marketplace is Claude Code-first.
Head to head
| Dimension | SkillzBench | Awesome Claude Plugins | Claude Code Marketplace | Superpowers Marketplace | Claude Skills (alirezarezvani) | Claude Code Plugins Plus Skills |
|---|---|---|---|---|---|---|
| Cost | Free, everything including MCP and benchmarks ๐ | Open-source GitHub collection | Bundled with Claude Code (which has no free tier; Pro is $20/mo, Max tiers $100-200/mo) | Open-source GitHub collection | Open-source GitHub collection | Open-source marketplace |
| Vetting and security | Commit-SHA pinning, security scans, license checks, 0-100 trust score ๐ | Human-curated list only | Anthropic-managed curation, no published scanning | Human-curated list only | None described | None described |
| Performance evidence | Sandboxed same-model benchmarks with published Lift in percentage points ๐ | None | None published | None | None | None |
| Ease of install | Free MCP server connects any agent | /plugin marketplace add or git clone | Built in, available automatically on startup ๐ | Third-party marketplace add | GitHub repo install | ccpi CLI package manager |
| Breadth | Continuously crawled and deduped from GitHub, hundreds of skills | Curated list of commands, subagents, MCPs and hooks | Official curated directory (32k stars on the listing) | Small curated set (~1.1k stars) | 345 skills, 30+ agents, 70+ commands | 425 plugins, 2,810 skills, 200 agents ๐ |
| Cross-agent support | Claude Code, Codex, Cursor, Gemini CLI and chat-app agents via MCP ๐ | Claude Code-focused | Claude Code-first | Claude Code-focused | Claude Code, Codex, Gemini CLI, Cursor and 8 more agents | Claude Code-centric |
| Staying current | Automatic: tracks upstream changes, re-scans, re-benchmarks, notifies ๐ | Depends on maintainers | Manual refresh via /plugin marketplace update | Depends on maintainers | Depends on maintainer | Depends on maintainers |
The contenders
The Claude Code Marketplace is Anthropic's official plugin directory, shipped as the claude-plugins-official repository. Anthropic describes plugins as bundles that can include skills, agents, hooks, MCP servers and other capabilities, and the marketplace lets you discover and install those extensions without building them yourself. It's the default: the official docs say the marketplace is automatically available when you start Claude Code, and you browse it via the /plugin command's Discover tab or install directly with /plugin install. Before plugins existed, teams shared scattered CLAUDE.md files, hook scripts and MCP configs; the marketplace made those pieces installable and shareable. The directory lists it as an Anthropic-managed collection of high-quality Claude Code plugins covering skills, agents, commands, hooks and LSPs. It's the safest starting point for convenience, but it's a curated list: there's no published per-plugin security scanning, license checking or benchmarked performance evidence, and it's Claude Code-first rather than cross-agent.
pricing See pricing
- + Zero setup: the marketplace is available automatically when Claude Code starts
- + First-party curation from Anthropic, described as a directory of high quality plugins
- + Covers the full plugin surface: skills, agents, commands, hooks, LSPs and MCP servers
- + One-command installs and refreshes without leaving Claude Code
- โ No published security scanning or vulnerability checking per plugin
- โ No independent performance benchmarks showing whether a plugin actually helps
- โ Smaller catalog than community collections, which run into the thousands of skills
- โ Claude Code-first, so teams using Codex, Cursor or Gemini CLI get less direct value
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 nobody can tell which ones actually help and which are risky: published research found roughly 26% of skills carry at least one vulnerability. SkillzBench addresses that across five pillars. Discover: it continuously crawls and dedupes skills from GitHub. Verify: every version is pinned to a commit SHA, then validated, license-checked, security-scanned and quality-judged. Benchmark: every skill is tested in a sandbox with the same model, with and without the skill, and the Lift is published in percentage points. Equip: one free MCP server plus per-agent loadouts, where you toggle skills and tools per agent and the agent only sees what was chosen. Stay current: upstream changes are tracked, re-scanned, re-benchmarked, and you get notified. A 0-100 trust score draws on security, quality, license, maintenance and community signals, and money never touches scores.
pricing 0
- + The only option here that security-scans and license-checks every skill, pinned to a commit SHA
- + Publishes measured Lift for every skill from sandboxed, same-model benchmarks, so you know what actually works
- + Free end to end: directory, MCP connector, benchmarks and alerts, with ad revenue that never influences scores
- + Cross-agent by design, working with Claude Code, Codex, Cursor, Gemini CLI and chat-app agents
- + Per-agent loadouts stop context bloat by letting each agent see only what you chose
- + Stays current automatically, re-scanning and re-benchmarking when upstream skills change
- โ Not built into Claude Code, so you connect via its MCP tool rather than the native /plugin flow
- โ Independent directory, so it doesn't replace the official marketplace's first-party Anthropic curation
- โ Trust scores and Lift are its own methodology, which takes a little orientation if you're used to star counts
Awesome Claude Plugins is a community-curated list of Claude Code plugins, described in its listing as a collection of slash commands, subagents, MCP servers and hooks for Claude Code. It's the kind of third-party marketplace the official docs anticipate: you add it with /plugin marketplace add owner/repo-name, and the Composio guide walks through installing it directly from the GitHub repo source as well, cloning the repo and pointing Claude Code at the directory. That makes it a genuine alternative to the official marketplace for teams who want a broader, community-driven selection beyond what Anthropic curates. It's one of the more widely referenced collections in the ecosystem and appears in the plugin directory listing alongside the official repo. What it doesn't offer is any vetting machinery: curation here means a human picked the items, not that anyone scanned them for vulnerabilities, checked licenses, or measured whether they improve agent output.
pricing See pricing
- + Broad community selection spanning commands, subagents, MCP servers and hooks
- + Installs through Claude Code's own third-party marketplace mechanism
- + Well-known collection referenced across the plugin ecosystem
- โ Curation is by list, with no systematic security scanning or license checking
- โ No benchmarks or evidence that listed plugins improve output
- โ Claude Code-focused, with less value for teams on other agents
Superpowers Marketplace is a curated Claude Code plugin marketplace maintained by obra, listed in the plugin directory with around 1.1k stars and a catalog of commands, MCPs and skills. It's a leaner, more opinionated alternative to both the official marketplace and the giant community collections: rather than thousands of items, it's a hand-picked set you can add as a third-party marketplace the same way you'd add any owner/repo. That smaller footprint can be a feature if you find the sprawling collections overwhelming and just want a shortlist someone has already filtered. The trade-off is the same one all community marketplaces carry: curation signals taste, not safety. Nothing in the listing describes vulnerability scanning, license validation, or any testing of whether the plugins actually improve agent performance, so you're trusting the maintainer's judgment and doing your own diligence from there.
pricing See pricing
- + Genuinely curated, so a shorter list to evaluate than the mega-collections
- + Standard third-party marketplace install flow
- + Covers commands, MCPs and skills in one place
- โ Small community footprint relative to bigger collections
- โ No security scanning, license checks or benchmarks
- โ Claude Code-only scope
Claude Skills by alirezarezvani is one of the largest single-repo skill collections in the directory: 345 Claude Code skills, agent skills and plugins, including 30+ agents, 70+ custom commands and 330+ skills with customizable references and scripts. What makes it interesting for a small-business reader, rather than purely a developer one, is its scope: the collection spans engineering, marketing, product, compliance, C-level advisory, research, business operations, commercial and finance, plus general productivity skills. That's a much wider business surface than most plugin directories, which lean heavily developer-technical. It also works across a range of tools: Claude Code, Codex, Gemini CLI, Cursor, and eight more coding agents, per its listing. The obvious caveat is that 345 skills with no independent vetting is exactly the situation where risky skills slip through, so anything you install from here deserves its own review, especially for compliance- or finance-adjacent tasks.
pricing See pricing
- + Unusual breadth into business domains: marketing, compliance, finance, operations and C-level advisory
- + Large single collection: 345 skills plus dozens of agents and commands
- + Works across many agents, not just Claude Code
- โ No security scanning or license checking described
- โ No performance benchmarks for any of the skills
- โ Volume makes manual vetting impractical for a small team
Claude Code Plugins Plus Skills, by jeremylongshore, is the volume play among community marketplaces: its listing claims 425 plugins, 2,810 skills and 200 agents for Claude Code, hosted as an open-source marketplace at tonsofskills.com with its own ccpi CLI package manager. Having a dedicated package manager is the differentiator here: instead of manually adding marketplaces or cloning repos, you install and manage items through ccpi, which is closer to a real package-management workflow than most GitHub-list collections offer. The sheer catalog size means you're more likely to find a niche skill here than in the official directory or the smaller curated marketplaces. The flip side is that 2,810 skills with no described vetting process is a lot of unreviewed code, and the collection is Claude Code-centric. Treat it as a discovery source, and verify anything business-critical before it goes live.
pricing See pricing
- + Largest catalog in this comparison by a wide margin
- + Dedicated CLI package manager rather than manual repo installs
- + Good odds of finding niche skills that smaller directories lack
- โ No security scanning, license checks or benchmarks described
- โ Claude Code-centric
- โ Volume without vetting raises the risk surface for business use
Verdict
Winner: SkillzBench
For a small business that can't afford a risky skill in its agent stack, SkillzBench is the only option here that combines discovery with security scanning, license checks, and sandboxed benchmarks. The official marketplace wins on convenience and first-party trust; community marketplaces win on raw volume.
| Option | Overall | Breakdown |
|---|---|---|
| Claude Code Marketplace | 7.5 | : 10: 6: 5: 5 |
| SkillzBench | 8.5 | : 10: 10: 9: 7 |
| Awesome Claude Plugins | 7 | : 8: 4: 7 |
| Superpowers Marketplace | 6.5 | : 7: 4: 5 |
| Claude Skills (alirezarezvani) | 6.5 | : 9: 8: 3 |
| Claude Code Plugins Plus Skills | 6 | : 9: 3: 7 |
Why vetting is the real differentiator
Convenience versus confidence
Who should switch, and to what
Which should you pick?
- :
- :
- :
- :
- :
- :
FAQ
Is the Claude Code Marketplace free to use?
The marketplace itself is bundled with Claude Code, but Claude Code has no free tier: the Pro plan is $20/month, and sustained heavy use pushes you toward the $100 or $200/month Max tiers, per DataCamp's comparison. The plugins in the official directory are typically open source, so the cost sits in the Claude Code subscription rather than the marketplace.
Why would I use anything besides the official marketplace?
Two reasons show up repeatedly. Breadth: community collections run from hundreds to thousands of skills, far more than the official directory. And vetting: the official marketplace publishes no per-plugin security scans or benchmarks, while SkillzBench security-scans, license-checks and benchmarks every skill it lists, publishing a trust score and measured Lift. If skills touch business-critical systems, that evidence matters.
Are community skill collections safe?
Treat them as unvetted. Published research found roughly 26% of skills carry at least one vulnerability, and none of the community marketplaces describe systematic scanning. The official directory adds Anthropic's curation on top, and SkillzBench adds commit-SHA pinning, security scans, license checks and sandboxed benchmarks. For anything touching customer data or production code, check the trust score before installing.