# Microsoft Agent Framework + Dapr

Make your Microsoft Agent Framework agents production-ready. These tutorials show you how to add fault tolerance using Dapr.

- [Durable Workflows](https://docs.diagrid.io/getting-started/quickstarts/ai-agents?agentframework=microsoft-dotnet#1-log-in-to-catalyst) — Wrap agent tasks in Dapr workflows. Automatic retries, checkpointing, and crash recovery.

---

## Why Dapr for Microsoft Agent Framework?

| Challenge | Dapr Solution |
|-----------|---------------|
| Agent crashes lose all progress | **Durable workflows** checkpoint every step |
| Agents tightly coupled, hard to scale | **Pub/sub** decouples communication |
| Chat memory lost on restart | **State store** persists agent memory across restarts |
| Custom retry logic everywhere | **Built-in retries** with exponential backoff |

:::note\[Conversation API required]
Unlike the third-party Python frameworks, the Microsoft Agent Framework integration ([`diagridio/dotnet-ai`](https://github.com/diagridio/dotnet-ai)) reaches its LLM through a Dapr [conversation component](https://docs.diagrid.io/references/catalyst/components-reference/conversation/openai) — register the conversation client with `AddDaprConversationClient()` and pass a `conversationComponentName` when you add an agent. A conversation component is **required** here, not optional. See the [framework requirements table](https://docs.diagrid.io/develop/agents#framework-requirements).
:::

**Dapr & AI University — Learn this hands-on:** [Make MAF agents reliable with Dapr Workflow](https://www.diagrid.io/university/ai-agents-maf?utm_source=docs&utm_medium=cta&utm_campaign=ai-agents-maf)

Add durable state and crash recovery to Microsoft Agent Framework agents with Dapr Workflow. Requires an OpenAI API key.

Intermediate · 30 min · .NET