Timsunzhuping

PluginWorld — AI skill for Claude Code

AI community

跨生态 AI 插件聚合市场 — 一个搜索框找到 dsh / Claude Code / MCP 的所有插件 · Plug in.

How to install PluginWorld

This entry records only its repository, not the path inside it, so there is no exact command to give. Open Timsunzhuping/PluginWorld and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What PluginWorld does

跨生态 AI 插件聚合市场 — 一个搜索框找到 dsh / Claude Code / MCP 的所有插件 · Plug in. The world is ready.

Alternatives in AI

AI Engineering Hub — In-depth tutorials on LLMs, RAGs and real-world AI agent applications 37.2k ★
Agentmemory — #1 Persistent memory for AI coding agents based on real-world benchmarks 1.9k ★
Dream Xi AI — 🏄 Dream XI AI turns isolated AI agents into a World Cup-level dream team.🏄 Dream XI AI 是一个面向复杂任务执行的 Multi-Ag 441 ★

README

PluginWorld

**The cross-ecosystem AI plugin marketplace** — one search box for every plugin across DeepSeek Harness (dsh), Claude Code and MCP, with unified quality and security scores.

Plug in. The world is ready. · www.pluginworld.ai

![ecosystems](https://img.shields.io/badge/ecosystems-dsh%20·%20claude--code%20·%20mcp-c6ff00?labelColor=16161a)

Architecture

Next.js (App Router) + TypeScript + Tailwind v4 — SSG/ISR, SEO-first
Dual-mode data layer
- No DATABASE_URL → snapshot mode: reads src/data/seed/, deployable with zero config
- With DATABASE_URL → Postgres mode (Supabase/Neon + Drizzle): FTS search, claims, events, trending
Daily community scan (scripts/sync/, GitHub Actions cron at 02:00 UTC) over the world's mainstream plugin communities:
- GitHub topics (dsh-plugin / claude-code-plugin / mcp-server, top 500 by stars each)
- Official MCP Registry (registry.modelcontextprotocol.io)
- npm registry (keywords dsh-plugin, cordis-plugin, claude-code, mcp-server, ranked by monthly downloads)
- Glama MCP directory (metaregistry)
Pre-index security scan → rating: every candidate is scanned BEFORE listing; grade D is quarantined, never indexed (see below)
Public REST API (/api/v1/*): built for AI agents, CORS open, 60 req/min rate limit

Security rating (assigned before indexing, re-checked daily)

The scanner (`src/lib/security.ts`) checks: malicious install-script patterns, code-obfuscation signals, typosquatting against popular names (edit-distance + adoption-gap heuristic), suspicious URLs (shorteners / raw IPs / punycode), leaked credentials in docs (also auto-redacted), insecure http:// MCP remotes, and star-velocity anomalies (star farming).