Chapter 57: Dapr Actors & Workflows

Add durable state and orchestration to your agents. This chapter builds a dapr-actors-workflows skill that uses Dapr virtual actors for per-entity state and Dapr Workflows for long-running, reliable processes.

---

Goals

• Understand the actor model: virtual actors, turn-based concurrency, lifecycle
• Implement Dapr Actors for agent state (sessions, conversations, tasks)
• Use timers and reminders for scheduled work
• Design Dapr Workflows for durable orchestration with retries/compensation
• Combine actors and workflows for complex agent behaviors
• Package patterns into a reusable skill

---

Lesson Progression

• Actor model foundations
• Dapr Actor fundamentals and state management
• Timers and reminders
• Workflow patterns: sequential, parallel, saga/compensation
• Failure handling and retries
• Capstone: stateful agent with actors + workflows; finalize the skill

Each lesson ends with a reflection to test, find gaps, and improve the skill.

---

Outcome & Method

You finish with a stateful Task API that uses actors for per-entity state and workflows for long-running tasks, plus a Dapr actors/workflows skill. The chapter follows the skill-first flow: learn, apply, capstone, finalize.

---

Prerequisites

• Chapters 49-53 (containerized, Kubernetes, Helm, Dapr Core)
• Chapter 55 observability for monitoring actors/workflows