dapr-workflows

Build durable, autonomous, resilient and observable AI agents with built-in workflow orchestration, security, statefulness and telemetry.

Design proposals for Dapr features: new building blocks and APIs for workflows, the scheduler, pub/sub, actors, the LLM Conversation API and MCP servers, reviewed in the open by the Dapr community before they ship.

Durable execution for AI agents in Python: make LangGraph, CrewAI, Pydantic AI, OpenAI Agents, Google ADK, Strands and more fault-tolerant with Dapr Workflows on Diagrid Catalyst. Crash-proof agents, no re-run LLM calls.

Durable execution for Microsoft Agent Framework agents in .NET, on Dapr Workflow and Diagrid Catalyst

Durable execution for AI agents and workflows: Claude Code plugin, agent skills and remote MCP server for building, running and debugging Dapr Workflows and Durable Agents on Diagrid Catalyst. Also works in Codex and GitHub Copilot.

An MCP Server to interact with the Dapr APIs

Durable execution for TypeScript AI agents: Mastra and more, as Dapr Workflows on Diagrid Catalyst

Durable execution for Java AI agents: make Spring AI ChatClient calls, tool calls and chat memory crash-proof with Dapr Workflows on Diagrid Catalyst. Spring Boot starter, zero code changes, on Maven Central.

Implement Radius Recipes as Dapr durable workflows in Go — schedule over REST, orchestrate activities, return recipe outputs (values, secrets, resources)

AI Pizza Shop - Dapr Workflow DACA Challenge

Go-kratos Coffee Shope Example

Durable execution for Go AI agents: LangChainGo and Eino graphs as Dapr Workflows on Diagrid Catalyst

Dapr workflow orchestration for infrastructure automation (golden image builds, provisioning, etc.)