A production-ready marketplace for hosting, managing, and securing AI agent skills.
- Skill Hosting - Upload and manage skill packages (ZIP files with SKILL.md)
- Security Scanning - Pattern-based + AI-powered security analysis
- Skill Evaluation - Automated test execution with queued processing
- Team Management - Collaborate on skills with team members
- Skill Bundles - Group related skills together
- File Preview - Browse skill contents before download
- Frontend: Next.js 14, React, Tailwind CSS, shadcn/ui
- Backend: Next.js API Routes, Prisma ORM
- Database: PostgreSQL
- Queue: BullMQ + Redis
- Storage: S3/MinIO
- AI: Anthropic API (Claude)
- Node.js 18+
- PostgreSQL
- Redis
- MinIO (or S3-compatible storage)
# Clone the repository
git clone https://github.com/YOUR_USERNAME/skillmarketplace.git
cd skillmarketplace
# Install dependencies
pnpm install
# Set up environment variables
cp .env.example .env
# Edit .env with your configuration
# Set up database
pnpm db:generate
pnpm db:push
pnpm db:seed
# Start development server
pnpm dev# Required
DATABASE_URL=postgresql://user:password@localhost:5432/skillmarketplace
NEXTAUTH_SECRET=your-secret-key
NEXTAUTH_URL=http://localhost:3000
REDIS_URL=redis://localhost:6379
# Storage (MinIO/S3)
S3_ENDPOINT=http://localhost:9000
S3_ACCESS_KEY=minioadmin
S3_SECRET_KEY=minioadmin123
S3_BUCKET=skillmarketplace
# AI Security Analysis (optional)
AI_SECURITY_ENABLED=true
AI_SECURITY_API_KEY=your-anthropic-api-key
AI_SECURITY_BASE_URL=https://api.anthropic.com
AI_SECURITY_MODEL=claude-sonnet-4-20250514# Start dev server
pnpm dev
# Run tests
pnpm test # Unit tests
pnpm test:watch # Watch mode
pnpm test:coverage # With coverage
pnpm test:e2e # E2E tests
# Database
npx prisma studio --port 5556 # Open DB GUI
pnpm db:seed # Seed test data
# Storage
pnpm minio:start # Start MinIO container
pnpm minio:console # Open MinIO consoleAfter running pnpm db:seed:
| Password | Role | |
|---|---|---|
| admin@example.com | password123 | Admin |
| alice@example.com | password123 | User |
src/
├── app/
│ ├── (auth)/ # Authentication pages
│ ├── (dashboard)/ # Protected dashboard
│ ├── (marketplace)/ # Public marketplace
│ ├── api/ # REST API routes
│ └── actions/ # Server actions
├── lib/
│ ├── security/ # Security scanning (pattern + AI)
│ ├── eval/ # Evaluation queue system
│ ├── specification/ # SKILL.md parsing
│ ├── storage/ # S3/MinIO storage
│ └── teams/ # Team management
└── components/ # React components
prisma/
├── schema.prisma # Database schema
└── seed.ts # Test data seeding
tests/
├── unit/ # Jest unit tests
├── integration/ # API integration tests
└── e2e/ # Playwright E2E tests
SkillHub provides two-layer security analysis:
-
Pattern Scanner - Regex-based detection of:
- Code injection (eval, Function)
- Command injection (exec, spawn)
- Hardcoded credentials
- Path traversal
- And more...
-
AI Analyzer - Semantic analysis using Claude:
- Context-aware vulnerability detection
- Permission mismatch detection
- Security recommendations
本项目使用 智谱 GLM Coding 提供的 AI 能力进行开发。
MIT