Skip to main content

What is Agent Forge?

Agent Forge is a modern TypeScript framework for building AI agent systems using a decorator-based architecture. It simplifies the creation of single agents, multi-agent teams, and distributed agent networks.

Key Features​

Decorator-Driven Development​

  • @agent - Define agents with configuration
  • @tool - Add capabilities to agents
  • @llmProvider - Configure LLM connections
  • @forge - Initialize your application

Multi-Agent Orchestration​

  • Teams - Hierarchical agent coordination
  • Workflows - Sequential agent execution
  • A2A Protocol - Distributed agent networks

Rich Ecosystem​

  • Built-in Tools - Web search, content extraction
  • MCP Integration - Model Context Protocol support
  • RAG Support - Knowledge base integration
  • Plugin System - Extensible architecture

Architecture Overview​

Agent Forge uses modern TypeScript decorators to provide a clean, declarative way to build agent systems:

@tool(WebSearchTool)
@agent({
  name: "Research Assistant",
  role: "Research Specialist", 
  description: "Finds and analyzes information",
  objective: "Provide accurate research results",
  model: "gpt-4",
  temperature: 0.4
})
class ResearchAgent extends Agent {}

@llmProvider("openai", { apiKey: process.env.OPENAI_API_KEY })
@forge()
class MyApplication {
  static forge: AgentForge;
  
  static async run() {
    const agent = new ResearchAgent();
    const result = await agent.run("Research the latest AI trends");
    console.log(result.output);
  }
}

await MyApplication.run();

Why Decorator-Based?​

The decorator approach provides several advantages:

  • Type Safety - Full TypeScript support with IntelliSense
  • Clean Code - Configuration separated from logic
  • Composition - Mix and match decorators as needed
  • Scalability - Easy to manage complex agent systems

Next Steps​

Ready to get started? Continue with: