AgenticGoKit Documentation

Welcome to AgenticGoKit — the production-ready Go framework for building intelligent multi-agent AI systems.

Python has LangChain. Go has AgenticGoKit.

Quick Links

Quick Start — Build your first agent in 5 minutes
Installation — Install and configure AgenticGoKit
Core Concepts — Understand agents, workflows, and tools
Streaming — Real-time responses and live UI updates
Workflows — Multi-agent orchestration patterns
Memory & RAG — Conversational memory and knowledge retrieval
Tool Integration — MCP tools and custom tool development

Why AgenticGoKit?

AgenticGoKit brings the power of AI agent frameworks to Go developers:

🚀 Production-Ready — Battle-tested APIs designed for reliability and scale
⚡ Go Performance — Leverage Go's speed, concurrency, and type safety
🔧 Builder Pattern — Fluent, composable agent configuration
🌊 Native Streaming — First-class streaming support for real-time UX
🔗 MCP Support — Model Context Protocol for tool integration
🧠 Memory & RAG — Built-in conversation history and knowledge retrieval

Overview

go

// Build an AI agent in 3 lines of code
agent, _ := agenticgokit.NewBuilder("assistant").
    WithLLM("openai", "gpt-4").
    Build()

result, _ := agent.Run(ctx, "What is Go?")
fmt.Println(result.Content)

This documentation covers all core features, advanced topics, and complete examples to help you build, deploy, and scale your AI agents.

Note: Looking for legacy APIs (core/vnext)? See the Legacy Documentation or Migration Guide.