The 5 Best Claude Code Marketplace Alternatives in 2026

Anthropic's official plugin marketplace for Claude Code, available as the claude-plugins-official repo. Per Anthropic's docs, 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. Plugins are bundles that can include skills, agents, hooks, MCP servers and other capabilities, so instead of sharing scattered CLAUDE.md files, hook scripts and MCP configs, teams package workflows and install them as needed. The directory is described by Anthropic as an official, managed collection of high-quality Claude Code plugins, and it covers the categories developers reach for most: LSP code-intelligence plugins (TypeScript, Pyright, Rust Analyzer, Go), MCP integrations (GitHub, Linear, Slack, Vercel, Sentry, Supabase), code review plugins, and security plugins like Security Guidance and Semgrep. Third-party marketplaces can be added alongside it with /plugin marketplace add owner/repo-name.

pricing 0

• + Zero setup: the official marketplace ships with Claude Code and refreshes with /plugin marketplace update
• + Anthropic-managed, so plugin quality and compatibility with Claude Code are maintained at the source
• + One-command installs from a trusted repo reduce the risk of pulling in broken or malicious packages
• + Covers the highest-value categories first: LSP code intelligence, MCP integrations, code review and security

• − Claude Code only; skills and plugins don't port to Codex, Cursor or Gemini CLI the way cross-agent collections do
• − Smaller catalog than community collections, which list hundreds to thousands of items
• − No published independent benchmarking or security scanning of listed plugins, so 'high quality' is the vendor's own judgment
• − Using it at all requires an Anthropic subscription, and heavy agent use pushes users toward $100-200/month Max tiers per DataCamp's pricing breakdown

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) are everywhere but nobody can tell which ones help and which are risky: published research found roughly 26% of skills carry at least one vulnerability. SkillzBench continuously crawls and dedupes skills from GitHub, pins every version to a commit SHA, validates and license-checks each one, security-scans it, and quality-judges it. It then benchmarks every skill in a sandbox with the same model, with and without the skill, and publishes the Lift in percentage points. A 0-100 trust score combines security, quality, license, maintenance and community signals, and Lift is always shown separately from ratings so money never touches scores. Per-agent loadouts let you toggle skills and tools per agent, so each agent only sees what you chose for it, and SkillzBench tracks upstream changes, re-scans, re-benchmarks and notifies you when things shift.

pricing 0

• + The only option here that independently security-scans and license-checks every skill version, pinned to a commit SHA
• + Sandbox benchmarking shows measured Lift in percentage points, so you know a skill actually helps rather than trusting a star count
• + Works across Claude Code, Codex, Cursor, Gemini CLI and chat-app agents via one free MCP server
• + Per-agent loadouts stop context bloat: each agent only sees the skills and tools you toggled for it
• + Everything is free, including alerts; advertising on blog articles funds the platform and never influences scores or listings

• − It's a directory and connector, not a plugin runtime: you still install and run skills inside your agent
• − As a newer independent platform it doesn't carry Anthropic's first-party integration the official marketplace has
• − Benchmark coverage depends on its continuous crawl, so brand-new skills may not have Lift results yet

https://skillzbench.com

Composio's awesome-claude-plugins is one of the most-used third-party marketplaces for Claude Code, installable via /plugin marketplace add or by cloning the repo and running claude --plugin-dir. It's a large, categorized list spanning slash commands, subagents, MCP servers and hooks, organized by what developers actually reach for: code-intelligence LSP plugins (TypeScript, Pyright, Rust Analyzer, Go), MCP integration plugins (GitHub, Linear, Slack, Vercel, Sentry, Supabase), code review plugins (Code Review, PR Review Toolkit, CodeRabbit, Code Simplifier), and security plugins (Security Guidance, Semgrep). Installing from the repo also sets up Composio Connect MCP, which lets Claude Code securely access apps like Jira, Linear and GitHub from a single MCP server. Composio pitches dynamic tool loading to avoid the context-window problem you hit when running many MCP servers at once, since each server loads all its tools into context.

pricing 0

• + Well-organized by use case, so you can find the right plugin type fast instead of scrolling one flat list
• + Ships a practical answer to the multi-MCP context problem via dynamic tool loading
• + Free and open source, with a one-line marketplace add inside Claude Code

• − Curated but not verified: no security scanning or benchmarking of listed plugins
• − The Composio Connect MCP funnels you toward Composio's commercial platform, which is a soft upsell
• − Claude Code-focused, so cross-agent teams get partial coverage at best

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

aitmpl.com maintains a browsable web directory of Claude Code plugin collections, skill registries and component marketplaces on GitHub, currently listing 34 collections. Each entry shows the collection's GitHub star count and tags (skills, agents, commands, hooks, MCPs, LSPs), which makes it a genuine discovery layer over the fragmented ecosystem: you can compare Anthropic's official claude-plugins-official repo (32k stars) against community heavyweights like Ecc by affaan-m (229k stars), Claude Mem by thedotmack (87k stars), Claude Hud by jarrodwatts, Compound Engineering by EveryInc, and HashiCorp's official agent skills. It also surfaces specialized collections you'd never find otherwise, like Timescale's Postgres MCP plugin, Airtable's official plugins, and niche tools like Cartographer, which maps codebases with parallel AI subagents. Because it's a web directory rather than an in-terminal marketplace, you browse first, then install the underlying repo from GitHub.

pricing 0

• + The fastest way to see the whole ecosystem at a glance, including collections you'd never search for directly
• + Star counts and tags give you a rough popularity and purpose signal before you click through
• + Free, no account, no install

• − Popularity is the only ranking signal; stars don't tell you a skill is safe or effective
• − It's a meta-directory: you still install from each underlying GitHub repo, so setup varies per collection
• − No verification, licensing checks or benchmarking of anything listed

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

One of the most popular single-repo skill collections, with 345 Claude Code skills, agent skills and plugins: 30+ agents, 70+ custom commands, 330+ skills and customizable references and scripts. What sets it apart from most alternatives is scope beyond engineering. The collection spans marketing, product, compliance, C-level advisory, research, business operations, commercial and finance skills alongside the usual developer tooling, which makes it one of the few marketplaces that speaks directly to non-developer business users. It also works across a wide agent footprint: Claude Code, Codex, Gemini CLI, Cursor and eight more coding agents, so a small business standardizing skills across different tools can pull from one repo. Like other community collections it's installable as a third-party marketplace, and its 22k stars put it among the most-used community repos in the aitmpl.com directory.

pricing 0

• + Rare breadth into business domains, not just developer workflows
• + One repo covers 13+ coding agents, so mixed-tool teams standardize in one place
• + Large and actively starred (22k), so dead-collection risk is lower than with niche repos

• − A single maintainer's bundle: quality varies item to item with no independent vetting
• − Volume makes it hard to know which of the 330+ skills are worth installing
• − No security scanning or benchmarking, which matters given research showing ~26% of skills carry vulnerabilities

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

The aptly named volume play: an open-source marketplace at tonsofskills.com listing 425 plugins, 2,810 skills and 200 agents for Claude Code, maintained by jeremylongshore and surfaced in the aitmpl.com directory. Its differentiator is tooling: it ships its own CLI package manager, ccpi, so you can install and manage items from the collection programmatically rather than copying files by hand. That makes it closer to a real package-registry experience than most GitHub-list alternatives. The sheer size means it's one of the few places you can reasonably expect to find a skill for an obscure workflow, and the plugin/skill/agent taxonomy keeps the 2,800-plus items navigable. The tradeoff is the flip side of volume: with no verification or benchmarking layer, you're trusting the maintainer's aggregation, and the collection's 2.5k stars suggest a smaller community reviewing it than the top-ranked repos.

pricing 0

• + Largest single collection found, so odds are good it has a skill for niche workflows
• + The ccpi CLI gives real package-management ergonomics instead of manual file copying
• + Free and open source

• − Volume-first with no verification layer; more items means more unvetted surface area
• − Smaller community (2.5k stars) reviewing it than the leading collections
• − Claude Code-centric despite the size

https://tonsofskills.com