airesponse

airesponse is a lightweight Go CLI tool for chatting with AI models from your terminal. It currently supports Google Gemini models, offers an optional REPL mode, and can persist conversations to YAML for later reuse.

Features

• 💬 One-shot prompts or interactive REPL sessions
• 🔁 Conversation history tracking
• 💾 Save and load conversations as YAML
• 🤖 Support for multiple Gemini models
• ⚡ Fast, dependency-light Go binary
• 🛠 Easy to extend with additional providers

Installation

Build from source

Requirements:

• Go 1.25.6
• A Google Gemini API token

git clone https://github.com/CodeIter/airesponse_golang.git
cd airesponse_golang.git
make build

The binary will be available at:

./bin/airesponse

Cross-compile

make cross-compile

This builds binaries for Linux, macOS, and Windows.

Configuration

Environment Variables

airesponse reads credentials from environment variables:

Variable	Description
GEMINI_API_TOKEN	Google AI API token (required for Gemini models)

Example:

export GEMINI_API_TOKEN="your-api-key-here"

Usage

Basic prompt

./bin/airesponse "Explain Go interfaces in simple terms"

If no prompt is provided, a default introductory prompt is used.

REPL mode (interactive chat)

./bin/airesponse --repl

Or shorthand:

./bin/airesponse -r

Exit the REPL with:

exit
quit
bye

Choose a model

./bin/airesponse --model gemini-flash-latest "Write a haiku about compilers"

Shorthand:

./bin/airesponse -m gemini-flash-latest "Write a haiku about compilers"

List supported models

./bin/airesponse --list

Save and load conversations

Use the --file flag to persist conversation history as YAML:

./bin/airesponse --repl --file chat.yaml

• If the file exists, it is loaded at startup
• If it does not exist, a new conversation is created
• The file is updated after every model response

This makes it easy to:

• Resume chats
• Inspect or edit conversations manually
• Feed prior context back into the model

License

MIT License. See the LICENSE file for details.