Agents

Agents are the core abstraction in ADK TypeScript. They encapsulate AI models, instructions, tools, and coordination logic to create autonomous programs that can understand instructions and complete complex tasks.

Agent Types

ADK TypeScript provides different types of agents for different use cases:

🤖 LLM Agents

Use language models for reasoning, decision-making, and tool usage

🔗 Sequential Agents

Execute sub-agents in order for predictable workflows

⚡ Parallel Agents

Run multiple agents simultaneously for faster processing

🔄 Loop Agents

Repeat agent execution until conditions are met

🏗️ Custom Agents

Build specialized agents with custom logic and behavior

👥 Multi-Agent Systems

Coordinate multiple agents for complex distributed tasks

Models & Configuration

🧠 Model Support

Configure and use different LLM models with your agents

Choosing the Right Agent Type

Agent Selection Guide

LLM Agents: Best for reasoning, conversation, and tool usage
Sequential Agents: Use for step-by-step workflows and pipelines
Parallel Agents: Ideal for independent tasks that can run simultaneously
Loop Agents: Perfect for iterative improvement and retry logic
Custom Agents: When you need specific behavior or integration patterns
Multi-Agent: For complex coordination and specialized task distribution

Getting Started

New to agents? Start with these foundational concepts:

LLM Agents - Learn the most common agent type
Workflow Agents - Understand orchestration patterns
Models - Configure LLM models and settings

Advanced Topics

Once you're comfortable with the basics:

Custom Agents - Build specialized agent types
Multi-Agent Systems - Coordinate multiple agents