@lov3kaizen/aigency-cli

@aigency/cli

Command-line interface for Aigency ADK - Build and orchestrate AI agents from your terminal.

Features

• 🚀 Quick Setup - Initialize with interactive prompts
• 💬 Interactive Chat - Chat with agents in your terminal
• 🤖 Agent Management - Create, list, and manage agents
• 🔌 Provider Support - Cloud and local providers
• 📦 Model Management - Pull and manage Ollama models
• ⚙️ Configuration - Persistent configuration management
• 🎨 Beautiful Output - Colored, formatted terminal output

Installation

# Global installation
npm install -g @aigency/cli

# Or use with npx
npx @aigency/cli init

Quick Start

1. Initialize

aigency init

This will guide you through:

• Choosing a provider (cloud or local)
• Configuring API keys or endpoints
• Creating a default agent

2. Start Chatting

aigency chat

Interactive chat session with your default agent.

3. Run One-Off Commands

aigency agent run default "What is the capital of France?"

Commands

aigencyinit

Initialize Aigency CLI configuration with interactive prompts.

aigencyinit

aigencychat

Start an interactive chat session.

aigencychat                        # Use default agent
aigencychat --agent my-agent       # Use specific agent
aigencychat --model llama3         # Override model

aigencyagent

Manage agents.

# Create a new agent
aigencyagent create

# List all agents
aigencyagent list

# Get agent details
aigencyagent get <name>

# Run an agent with a message
aigencyagent run <name> "Your message"

# Set default agent
aigencyagent default <name>

# Delete an agent
aigencyagent delete <name>

aigencyprovider

Manage providers.

# List all providers
aigencyprovider list

# Get provider details
aigencyprovider get <name>

# Add a new provider
aigencyprovider add

# Set default provider
aigencyprovider default <name>

# Delete a provider
aigencyprovider delete <name>

aigencymodel

Manage models (Ollama only).

# List available models
aigencymodel list

# Pull a model from Ollama
aigencymodel pull llama2

# Show popular models
aigencymodel popular

aigencyconfig

Show current configuration.

aigencyconfig

Examples

Cloud Provider (Anthropic)

# Initialize with Anthropic
aigencyinit
> Cloud Provider
> Anthropic
> [Enter API Key]

# Chat with Claude
aigencychat

Local Provider (Ollama)

# Initialize with Ollama
aigencyinit
> Local Provider
> Ollama
> http://localhost:11434

# Pull a model
aigencymodel pull llama2

# Chat with local model
aigencychat

Multiple Agents

# Create a coding assistant
aigencyagent create
> Name: coder
> Model: codellama
> System Prompt: You are a coding assistant...

# Create a writer assistant
aigencyagent create
> Name: writer
> Model: llama2
> System Prompt: You are a creative writer...

# Use specific agent
aigencychat --agent coder

Configuration

Configuration is stored in:

• Linux: ~/.config/aigency-cli/config.json
• macOS: ~/Library/Preferences/aigency-cli/config.json
• Windows: %APPDATA%\aigency-cli\config.json

Configuration Structure

{
  "defaultProvider": "anthropic",
  "defaultAgent": "default",
  "providers": {
    "anthropic": {
      "name": "anthropic",
      "type": "anthropic",
      "apiKey": "sk-ant-...",
      "timeout": 60000
    },
    "ollama": {
      "name": "ollama",
      "type": "ollama",
      "baseUrl": "http://localhost:11434",
      "timeout": 60000
    }
  },
  "agents": {
    "default": {
      "name": "default",
      "description": "Default agent",
      "model": "claude-sonnet-4-20250514",
      "provider": "anthropic",
      "systemPrompt": "You are a helpful assistant.",
      "temperature": 0.7,
      "maxTokens": 2048
    }
  }
}

Environment Variables

API keys can also be set via environment variables:

export ANTHROPIC_API_KEY=your_key_here
export OPENAI_API_KEY=your_key_here
export GEMINI_API_KEY=your_key_here

Supported Providers

Cloud Providers

• Anthropic - Claude models
• OpenAI - GPT models
• Google - Gemini models

Local Providers

• Ollama - Local LLM runtime
• LM Studio - GUI for local models
• LocalAI - OpenAI-compatible local API

Ollama Integration

The CLI has first-class support for Ollama:

# Make sure Ollama is running
ollama serve

# Pull popular models
aigencymodel pull llama2
aigencymodel pull mistral
aigencymodel pull codellama

# List available models
aigencymodel list

# Show popular models
aigencymodel popular

# Chat with local model
aigencychat

Tips & Tricks

1. Quick Chat

Create an alias for quick access:

alias bc="aigencychat"

2. Multiple Providers

Set up multiple providers for different use cases:

aigencyprovider add
> Name: anthropic-prod
> Type: Anthropic

aigencyprovider add
> Name: ollama-dev
> Type: Ollama

3. Specialized Agents

Create agents for specific tasks:

# Coding agent
aigencyagent create
> Name: code
> Model: codellama
> System Prompt: You are an expert programmer...

# Writing agent
aigencyagent create
> Name: write
> Model: llama2
> System Prompt: You are a creative writer...

# Use them
aigencychat --agent code
aigencychat --agent write

4. Verbose Mode

Get detailed output:

aigencyagent run default "Hello" --verbose

Troubleshooting

"No providers configured"

Run aigencyinit to set up your first provider.

"Provider not found"

List providers: aigencyprovider list

Add provider: aigencyprovider add

"Agent not found"

List agents: aigencyagent list

Create agent: aigencyagent create

Ollama Connection Error

Make sure Ollama is running:

ollama serve

Check the base URL in your provider configuration:

aigencyprovider get ollama

Model Not Found (Ollama)

Pull the model first:

aigencymodel pull llama2

Development

# Install dependencies
pnpm install

# Build the CLI
pnpm build

# Link for local testing
pnpm link --global

# Test commands
aigency--help

License

MIT License - see LICENSE for details

See Also

• @aigency/core - Core Aigency ADK
• @aigency/nestjs - NestJS integration
• Aigency Documentation

Built with ❤️ by lovekaizen