CLI AI Agent in Go

This project demonstrates how to build a command-line AI agent using Go and the OpenRouter API.

---

Setup

1. Initialize Go module

go mod init your_project_name

2. Create .env file

vim .env

Add your OpenRouter API key:

OPENROUTER_API_KEY=your_api_key_here

3. Install OpenAI Go SDK

go get github.com/openai/openai-go/v2

4. Create main.go

vim main.go

• Define OpenRouter API URL.
• Load API key from .env and check if defined.
• Create LLM client using OpenAI SDK with base URL and API key.
• Create a messages array and append your query using openai.UserMessage.
• Choose a model and define chat completion parameters.
• Create a ctx and call client.ChatCompletions.New(ctx, params).
• Print the AI response.

5. Export environment variables

export $(cat .env | xargs)

6. Run

go run main.go

---

Version 2: Interactive Chat

Enhance the CLI to wait for user input in a loop:

• Create a scanner:

scanner := bufio.NewScanner(os.Stdin)

• Print a prompt symbol:

fmt.Print(">")

• Read input:

if !scanner.Scan() { break }
input := strings.TrimSpace(scanner.Text())
if input == "" { continue }

• Append input to messages array.
• Send chat completion request.
• Print AI response.
• Loop back for the next input.

---

Version 3: Memory and Colors

Enhance further with:

• Append AI response to messages.
• Print message count.
• Limit messages array to last 4 messages if it grows too large.
• Add colors with fatih/color:

go get github.com/fatih/color

• Use colors to differentiate:
  • User queries
  • AI responses
  • Other info

This improves readability and visual distinction in the chat.

---

Run

go run main.go

Enjoy your interactive CLI AI agent!