Introduction

AgentKit is a Go toolkit designed to help developers build LLM agents and AI-powered applications.

Integrating large language models (LLMs) into Go applications is complicated and heavily dependent on the specific model provider you use. AgentKit standardizes LLM integrations across supported providers, enabling developers to focus on building great AI applications rather than dealing with technical complexity.

Why use AgentKit?

Building LLM agents requires handling multiple concerns:

Model Provider Abstraction: Different APIs for OpenAI, Anthropic, Google, etc.
Conversation Management: Message history, context windows, and memory
Tool Integration: Function calling and external system integration
Streaming Support: Real-time response handling
Error Handling: Robust failure recovery and retry logic

AgentKit handles all of this for you with a simple, unified Go API.

Quick Example

Here's how easy it is to create an agent with AgentKit:

package main

import (
    "context"
    "fmt"
    "log"
    "os"

    "github.com/model-box/agent-kit/agent"
    "github.com/model-box/agent-kit/model"
    "github.com/model-box/agent-kit/session"
)

func main() {
    // Create a model using chainable API
    model := model.Model("gpt-4o").
        SetAPIKey(os.Getenv("OPENAI_API_KEY"))
    
    // Create an agent using chainable API
    ag := agent.New().
        SetModel(model).
        SetSystemPrompt("You are a helpful assistant.")
    }

    // Create session and run
    ctx := context.Background()
    session := session.New(ag)
    result, err := session.Run(ctx, []agent.ChatMessage{agent.NewUserMessage("Hello! What can you help me with?")}, nil)
    if err != nil {
        log.Fatal(err)
    }

    fmt.Println(result.FinalOutput)
}