Skip to main content

Dapr Agents Python SDK

Build LLM-powered autonomous agentic applications using Dapr Agents with Catalyst. Dapr Agents is a framework for creating AI agents that can execute tasks, make decisions, and collaborate through workflows, while leveraging Dapr's distributed systems capabilities for reliable execution at scale.

Resources​

Prerequisites:

Installation​

Install Dapr Agents via pip:

# Install Dapr Agents
pip install dapr-agents

# Install Dapr Python SDK (required dependency)
pip install dapr

Initialize Dapr locally for development:

dapr init

Core Concepts​

Learn the fundamental concepts of building agentic applications with Dapr Agents.

Introduction

Overview of Dapr Agents and its key features for building LLM-powered autonomous agents

Why Dapr Agents

Understand the benefits and use cases for building agentic systems with Dapr

Core Concepts

Learn about agents, LLM integration, prompts, structured outputs, MCP support, and memory

Building Agents​

Create and configure agents with tools, memory, and workflows.

Getting Started

Install Dapr Agents and create your first agent with step-by-step instructions

Agentic Patterns

Common design patterns and use cases for building agentic systems

Integrations

Various integrations and tools available in Dapr Agents

Dapr Building Blocks​

Dapr Agents leverages Dapr's distributed systems capabilities for reliable agent execution.

Workflow

Build durable, long-running agent workflows with state management

State Management

Persist agent memory, conversation history, and execution state

Pub/Sub

Enable agent-to-agent communication and event-driven workflows

Service Invocation

Invoke external services and APIs from your agents

Getting Started​

Dapr Agents Quickstarts

Explore official Dapr Agents quickstart examples

Dapr Agents Documentation

Comprehensive Dapr Agents guide and API reference

Catalyst Quickstarts​

Durable Agent Quickstart

Build a durable travel assistant agent with conversation memory and state persistence

Multi-Agent Quickstart

Create multiple agents that collaborate through workflows