The 5 Best Claude Code Marketplace Alternatives in 2026

The Claude Code Marketplace is Anthropic's official directory for Claude Code plugins, managed under the anthropics/claude-plugins-official repository with roughly 32k stars. Plugins are packaged bundles that can include skills, agents, hooks, MCP servers, and LSP servers, replacing the old scattered mix of CLAUDE.md files, hook scripts, and custom slash commands. The marketplace is available automatically when you start Claude Code: run /plugin, browse the Discover tab, and install with a single command, refreshing via /plugin marketplace update if something is missing. The catalog leans developer-first, with strong coverage of code intelligence (TypeScript, Pyright, Rust Analyzer, Go LSP plugins), MCP integrations (GitHub, Linear, Slack, Vercel, Sentry, Supabase), code review packs, and security plugins like Security Guidance and Semgrep. Anthropic also maintains a separate Knowledge Work Plugins repository aimed at knowledge workers in Claude Cowork. The trade-off is scope: it serves Claude Code only, and it doesn't publish independent security scans or before/after benchmarks for what it lists.

pricing See pricing

• + Official and Anthropic-managed, so it's the most trustworthy default for Claude Code users
• + Zero setup: the marketplace is available automatically when Claude Code starts
• + Strong developer catalog: LSP, MCP, code review, and security plugins
• + Free to browse and install from

• − Serves Claude Code only; teams running Codex, Cursor, or Gemini CLI need something else
• − No published independent security scans or benchmark results for listed plugins
• − Smaller catalog than large community collections like AEcc or Claude Skills
• − Underlying Claude Code costs add up: no free tier, $20/mo Pro, $100-200/mo Max for heavy agent use

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, and nobody can tell which ones actually help and which are risky; published research found roughly 26% of skills carry at least one vulnerability. SkillzBench runs five pillars today: Discover (continuously crawling and deduping skills from GitHub), Verify (pinning every version to a commit SHA, validating, license-checking, security-scanning, and quality-judging), Benchmark (testing every skill in a sandbox with the same model, with and without the skill, and publishing the Lift in percentage points), Equip (one free MCP server plus per-agent loadouts, so the agent only sees what you toggled on), and Stay current (tracking upstream changes, re-scanning, re-benchmarking, and notifying you). A 0-100 trust score combines security, quality, license, maintenance, and community signals, and Lift is always shown separately. Everything is free, funded by ads around the blog; ad money never touches scores or listings.

pricing See pricing

• + Everything is free: directory, MCP connector, benchmarks, and alerts
• + Independent verification layer the official marketplace and community collections don't offer, addressing the ~26% vulnerability finding
• + Agent-agnostic: one MCP server serves Claude Code, Codex, Cursor, Gemini CLI, and chat-app agents
• + Lift benchmarks show measurable impact, always separate from ratings and money
• + Upstream change tracking keeps skills current instead of going stale

• − It's a separate site and MCP server rather than built into your editor's plugin command
• − Focused on skills rather than full plugin bundles with hooks and LSP servers
• − Smaller public profile than the official marketplace or the biggest community repos

https://skillzbench.com

AEcc, by affaan-m, is the most-starred community collection in the Claude Code plugin ecosystem at roughly 229k GitHub stars. It bills itself as an agent harness performance optimization system, packaging skills, instincts, memory, security, and research-first development practices for Claude Code, Codex, OpenCode, Cursor, and beyond. Where the official marketplace focuses on discrete plugins, AEcc is more of an opinionated operating system for your agents: it bundles memory and context management alongside skills and commands. That makes it attractive if you want a whole workflow in one install rather than assembling pieces yourself. The trade-offs are the usual ones for community collections: installation goes through the /plugin marketplace add flow from a GitHub repo, there's no independent security scanning or benchmark data published, and you're trusting one maintainer's curation even with very strong community adoption.

pricing See pricing

• + Huge community adoption (229k stars) signals active maintenance and real-world use
• + Covers memory and context persistence, which most collections skip
• + Works across Claude Code, Codex, OpenCode, and Cursor
• + Free and open on GitHub

• − No published security scans or benchmark results
• − Opinionated all-in-one design may conflict with your existing setup
• − Single-maintainer curation

https://github.com/affaan-m

Claude Skills by alirezarezvani is a broad, single-maintainer collection at roughly 22k GitHub stars: 345 skills, 30+ agents, and 70+ custom commands, plus around 330 skills with customizable references and scripts. Its distinguishing feature is audience breadth. Alongside engineering skills, it covers marketing, product, compliance, C-level advisory, research, business operations, commercial and finance, and daily productivity, which makes it one of the few collections genuinely relevant to non-developer small-business roles. It targets Claude Code, Codex, Gemini CLI, Cursor, and 8 more coding agents, so a marketing team and an engineering team can pull from the same repo. The caveats: it's a community collection with no independent security scanning, quality varies across 345 entries, and you'll spend some time sorting the useful from the filler.

pricing See pricing

• + Rare non-engineering coverage: marketing, compliance, finance, and operations skills
• + Very broad agent compatibility (11+ coding agents)
• + Large catalog with active star count
• + Free and open

• − No independent verification, security scanning, or benchmarks
• − Quality consistency is hard to guarantee across 345 skills from one maintainer
• − Browsing and selecting from such a large catalog takes effort

https://github.com/alirezarezvani

Claude Code Plugins Plus Skills, by jeremylongshore, is an open-source marketplace with roughly 2.5k GitHub stars listing 425 plugins, 2,810 skills, and 200 agents for Claude Code. Two things set it apart from a plain GitHub list: it has its own website at tonsofskills.com and it ships a CLI package manager called ccpi, so installing and managing that large catalog feels more like using a package registry than cloning a repo. It's Claude Code-focused, which keeps the scope tight but excludes teams running Codex, Cursor, or Gemini CLI. Like the other community marketplaces, there's no published independent security scanning or benchmarking, so with 2,810 skills the vetting burden falls on you. If you're a Claude Code power user who wants volume and a real package manager, it fits; if you want curation and verification, look elsewhere.

pricing See pricing

• + Largest raw skill count in this comparison (2,810 skills)
• + Real CLI package manager makes bulk management practical
• + Own website makes browsing easier than raw GitHub lists
• + Free and open-source

• − Claude Code only
• − No independent security or quality verification across such a large catalog
• − Smaller community (2.5k stars) than the biggest collections

https://tonsofskills.com

Buildwithclaude, by davepoon, is a single hub (roughly 3.2k GitHub stars) for finding Claude Skills, Agents, Commands, Hooks, Plugins, and Marketplace collections. Its value is orientation: rather than being one more collection, it indexes the ecosystem, including Claude Desktop, the Agent SDK, and OpenClaw alongside Claude Code. If you're new to the plugin and skills ecosystem and unsure whether you need a skill, a plugin, an MCP server, or a whole marketplace, Buildwithclaude is a sensible first stop to map the territory before committing. It's free and lightweight, but it's a directory of directories in spirit, so it doesn't add verification, benchmarking, or a package manager, and its catalog depth depends on the collections it links to.

pricing See pricing

• + Best starting point for understanding the ecosystem's moving parts
• + Covers Claude Desktop and Agent SDK, which most collections ignore
• + Free and simple

• − More of an index than a curated catalog
• − No verification, benchmarking, or package management
• − Smaller community than the major collections

https://github.com/davepoon

Composio approaches the same problem from the integration side. Its Composio Connect MCP lets Claude Code securely access apps like Jira, Linear, GitHub, and more from a single MCP server, installed with one command from the awesome-claude-plugins repo. The reason it earns a spot on an alternatives list is a real pain point with the marketplace's MCP plugins: install Linear, GitHub, Supabase, and Vercel MCPs together and they load all available tools into the context window, choking it as usage grows. Composio solves this with dynamic tool loading (tools load as needed based on context), a remote workbench for chaining multiple tools into complex workflows, and a bash tool for handling large files outside the context window. You can get started at dashboard.composio.dev within about a minute, per the source. It complements rather than replaces a skills directory: it's about connecting agents to your apps, not about finding and vetting skills.

pricing See pricing

• + Directly fixes the context-window problem of stacking multiple MCP servers
• + One install covers many app integrations
• + Fast setup, about a minute at dashboard.composio.dev

• − Focused on app integrations, not a skills catalog
• − No published independent benchmarking of skill quality
• − Another vendor account to manage alongside your agent tooling

https://composio.dev