Agents are the core building blocks of Agentuity. Each agent handles a specific task with schema validation, AI integration, and lifecycle hooks.
Agents vs Routes
When to create an agent vs handling requests directly
Creating Agents
Learn how to create agents with schemas and handlers
Workbench
Test agents locally with the built-in UI
Schema Libraries
Use Zod, Valibot, or the built-in schema library
AI Gateway
Route LLM calls through OpenAI, Anthropic, and more
AI SDK Integration
Use Vercel AI SDK with Agentuity
Streaming Responses
Return real-time streaming responses
State Management
Persist data with thread and session state
Calling Other Agents
Build multi-agent systems
Standalone Execution
Run agents outside HTTP requests
Evaluations
Automated quality checks after each execution
Events & Lifecycle
Hook into agent lifecycle events