Feature Map
Decorator-Driven Development
Build servers with decorators — @FrontMcp, @Tool, @Resource, @Prompt, and more
Dependency Injection
@Provider with scopes, this.get(), and hierarchical DI containers
Multi-App Composition
Compose multiple @App modules into a single server with isolated auth
Authentication & Authorization
Three-tier auth: public, transparent, and orchestrated OAuth 2.1
Deployment Targets
Deploy to Node.js/Docker, Vercel, AWS Lambda, or Cloudflare Workers
Plugin System
Extend your server with plugins, adapters, and context extensions
Testing Framework
E2E fixtures, custom matchers, interceptors, and HTTP mocking
Skill-Based Workflows
Teach AI multi-step workflows with @Skill and Agent Skills spec support
Jobs
Define typed jobs with retries, timeouts, and background execution
Workflows
Compose jobs into DAG-based multi-step pipelines
Architecture at a Glance
At-a-Glance Comparison
| Feature | What It Does | Key Decorator | Docs |
|---|---|---|---|
| Tools | AI-callable functions with typed I/O | @Tool | Tools |
| Resources | Read-only data endpoints | @Resource | Resources |
| Prompts | Reusable message templates | @Prompt | Prompts |
| Skills | Multi-step workflow guides | @Skill | Skills |
| Agents | LLM-powered autonomous actors | @Agent | Agents |
| Providers | Dependency injection services | @Provider | Providers |
| Plugins | Cross-cutting extensions | @Plugin | Plugins |
| Adapters | Dynamic capability generation | @Adapter | Adapters |
| Jobs | Reliable executable units of work | @Job | Jobs |
| Workflows | DAG-based multi-step pipelines | @Workflow | Workflows |
| Flows | Request lifecycle pipelines | @Flow | Flows |
| Apps | Modular capability containers | @App | Apps |