Skip to main content

​
Overview

The ADK CLI provides seven official starter templates, each designed for specific use cases. All templates are production-ready and include complete TypeScript configurations, example agents, and integration setups.

​
Selecting a Template

When creating a new project, you’ll be prompted to choose a template: 
adk new my-project
Or specify the template directly: 
adk new my-project --template simple-agent

​
Available Templates

​
🤖 Simple Agent

A minimal starter template for building general-purpose AI agents. Template ID: simple-agent Use Cases:
  • Learning ADK basics
  • General-purpose chatbots
  • Custom agent prototypes
  • CLI-based assistants
What’s Included:
  • Basic agent configuration
  • Example conversation flow
  • TypeScript setup
  • Development scripts
Project Structure: 
my-agent/
├── src/
│   └── index.ts          # Agent definition and entry point
├── agents/
│   └── agent.ts          # Exportable agent for CLI testing
├── package.json
├── tsconfig.json
└── .env.example          # Environment variables template
Create this template: 
adk new my-agent --template simple-agent
Example agent.ts: 
import { AgentBuilder } from '@iqai/adk';

const agent = new AgentBuilder()
  .withName('Simple Assistant')
  .withModel('gpt-4')
  .withInstruction(
    'You are a helpful assistant that answers questions clearly and concisely.'
  )
  .buildLlm();

export default agent;

​
🎮 Discord Bot

Integrate your AI agent with Discord for interactive server experiences. Template ID: discord-bot Use Cases:
  • Discord community assistants
  • Gaming help bots
  • Server moderation with AI
  • Interactive Discord experiences
What’s Included:
  • Discord.js integration
  • Message handling
  • Agent conversation flow
  • Bot configuration
Key Features:
  • Slash command support
  • Message listeners
  • Channel-based conversations
  • Role-based permissions
Create this template: 
adk new discord-bot --template discord-bot
Configuration: 
DISCORD_BOT_TOKEN=your_bot_token_here
DISCORD_CLIENT_ID=your_client_id_here
OPENAI_API_KEY=your_openai_key_here
Usage: 
npm run dev
The bot will connect to Discord and respond to messages mentioning it.

​
📱 Telegram Bot

Build AI-powered Telegram bots with seamless message handling. Template ID: telegram-bot Use Cases:
  • Telegram assistants
  • Customer support bots
  • Information retrieval bots
  • Interactive Telegram experiences
What’s Included:
  • Telegraf integration
  • Command handlers
  • Conversation state management
  • Bot configuration
Key Features:
  • Command support (/start, /help, etc.)
  • Inline keyboards
  • Photo and file handling
  • Group chat support
Create this template: 
adk new telegram-bot --template telegram-bot
Configuration: 
TELEGRAM_BOT_TOKEN=your_bot_token_here
OPENAI_API_KEY=your_openai_key_here
Usage: 
npm run dev

​
🚀 Hono Server

Build high-performance web APIs with Hono and AI agent endpoints. Template ID: hono-server Use Cases:
  • REST APIs with AI capabilities
  • Serverless agent deployments
  • Edge runtime agents
  • High-performance backends
What’s Included:
  • Hono web framework setup
  • RESTful agent endpoints
  • CORS configuration
  • Request validation
Key Features:
  • Ultra-fast routing
  • Middleware support
  • Edge runtime compatible
  • TypeScript-first
Create this template: 
adk new api-server --template hono-server
API Endpoints: 
POST /api/chat
GET /api/health
Example request: 
curl -X POST http://localhost:3000/api/chat \
  -H "Content-Type: application/json" \
  -d '{"message": "Hello, agent!"}'
Deploy to:
  • Cloudflare Workers
  • Deno Deploy
  • Vercel Edge Functions
  • AWS Lambda

​
🔌 MCP Starter

Integrate with Model Context Protocol (MCP) servers for enhanced agent capabilities. Template ID: mcp-starter Use Cases:
  • Agents with external tool access
  • Integration with MCP ecosystems
  • Enhanced agent capabilities
  • IDE and editor integrations
What’s Included:
  • MCP client setup
  • Server connection handling
  • Tool integration examples
  • Configuration templates
Key Features:
  • Connect to MCP servers
  • Access external tools
  • Extend agent capabilities
  • Resource management
Create this template: 
adk new mcp-agent --template mcp-starter
Configuration: 
import { AgentBuilder } from '@iqai/adk';
import { McpTools } from '@iqai/adk/mcp';

const tools = await McpTools.fromRemoteServer({
  command: 'npx',
  args: ['-y', '@modelcontextprotocol/server-filesystem'],
});

const agent = new AgentBuilder()
  .withName('MCP Agent')
  .withModel('gpt-4')
  .withTools(tools)
  .buildLlm();
Compatible MCP Servers:
  • Filesystem
  • Database
  • HTTP APIs
  • Custom servers

​
🌓 Near Shade Agent

Build agents that interact with the NEAR blockchain using Shade. Template ID: shade-agent Use Cases:
  • Blockchain interaction agents
  • DeFi assistants
  • NFT management bots
  • Web3 automation
What’s Included:
  • Near Shade integration
  • Wallet connection
  • Transaction handling
  • Blockchain queries
Key Features:
  • NEAR Protocol integration
  • Smart contract interaction
  • Token transfers
  • Account management
Create this template: 
adk new near-agent --template shade-agent
Configuration: 
NEAR_NETWORK=testnet
NEAR_ACCOUNT_ID=your-account.testnet
NEAR_PRIVATE_KEY=your_private_key_here
OPENAI_API_KEY=your_openai_key_here
Example operations:
  • Check wallet balance
  • Transfer NEAR tokens
  • Query smart contracts
  • Deploy contracts

​
⚡ Next.js Starter

Full-stack AI agent application with Next.js and Tailwind CSS. Template ID: next-js-starter Use Cases:
  • Full-stack AI applications
  • Customer-facing chat interfaces
  • SaaS products with AI
  • Interactive web experiences
What’s Included:
  • Next.js 14+ with App Router
  • Tailwind CSS styling
  • API routes for agents
  • Modern UI components
  • Responsive design
Key Features:
  • Server and client components
  • Streaming responses
  • Beautiful chat UI
  • SEO optimized
  • Production ready
Create this template: 
adk new ai-app --template next-js-starter
Project Structure: 
ai-app/
├── app/
│   ├── api/
│   │   └── chat/
│   │       └── route.ts       # Agent API endpoint
│   ├── page.tsx               # Home page with chat UI
│   └── layout.tsx             # Root layout
├── agents/
│   └── agent.ts               # Agent definition
├── components/
│   ├── chat.tsx               # Chat component
│   └── message.tsx            # Message component
├── public/                     # Static assets
├── styles/
│   └── globals.css            # Global styles
└── package.json
Development: 
npm run dev
Navigate to http://localhost:3000 to see your AI chat interface. Deploy to:
  • Vercel (recommended)
  • Netlify
  • AWS Amplify
  • Self-hosted

​
Template Comparison

TemplateComplexityUse CaseIntegration
Simple AgentBeginnerGeneral purposeNone
Discord BotIntermediateDiscord serversDiscord.js
Telegram BotIntermediateTelegram chatsTelegraf
Hono ServerIntermediateREST APIsHono
MCP StarterAdvancedTool integrationMCP
Shade AgentAdvancedBlockchainNEAR
Next.js StarterAdvancedFull-stack webNext.js

​
Template Features Matrix

FeatureSimpleDiscordTelegramHonoMCPShadeNext.js
TypeScript✅✅✅✅✅✅✅
CLI Testing✅✅✅✅✅✅✅
Web UI❌❌❌❌❌❌✅
External Integration❌✅✅✅✅✅❌
Production Ready✅✅✅✅✅✅✅
Hot Reload Support✅✅✅✅✅✅✅

​
Customizing Templates

All templates are fully customizable. After creation:
1
Modify agent configuration
2
Edit agents/agent.ts to change your agent’s behavior, model, or tools.
3
Add custom tools
4
Create new tools in src/tools/ and register them with your agent.
5
Update environment variables
6
Configure API keys and settings in .env.
7
Extend functionality
8
Add new files and features following TypeScript best practices.

​
Common Template Structure

All templates share a similar structure: 
project-name/
├── agents/              # Agent definitions for CLI
│   └── agent.ts
├── src/                 # Source code
│   ├── index.ts        # Main entry point
│   ├── tools/          # Custom tools (optional)
│   └── config/         # Configuration (optional)
├── .env.example         # Environment template
├── package.json         # Dependencies and scripts
├── tsconfig.json        # TypeScript configuration
└── README.md            # Template documentation

​
Testing Templates

All templates support testing with the CLI: Terminal chat: 
cd my-project
adk run
Web interface: 
cd my-project
adk web
Production start: 
npm run build
npm start

​
Next Steps

CLI Commands

Learn all CLI commands for testing and development

Agent Builder

Build custom agents with the AgentBuilder API

Tools System

Add tools to extend agent capabilities

Examples

Explore complete example implementations

Build docs developers (and LLMs) love

Get started for freeTalk to us