Pydantic AI agents connect to Azure Durable Task
Understand this story
Pydantic AI Harness releases the AzureDurableTaskDurability capability, connecting the agent loop to Microsoft Azure Durable Task so agents run durably on Azure Durable Task. Model requests, function tool calls, MCP calls, and durability operations each run as an activity by default; when not in an orchestration, agent behavior is the same as before. A simulator is available for local development, with the SDK listening on port 8080, the task hub named default, and the dashboard at localhost:8082; an Azure subscription is not required for now. The Harness is still at 0.x, and the API may change between minor versions.
AI overview based on source reports · Updated 39 min ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Pydantic AI ArticlesFeaturedDurable Pydantic AI agents with Azure Durable Task
Pydantic AI Harness now includes AzureDurableTaskDurability, a capability that runs the agent loop on Microsoft's Durable Task engine, recording each model request, tool call, MCP call, and durable operation as an activity so a replacement worker can replay history and resume after a crash. The post walks through a support agent example on the local DTS emulator, shows a worker killed mid-refund resuming without repeating completed activities, and covers replay rules (at-least-once activities, idempotency keys from ctx.run_id and ctx.tool_call_id, deterministic orchestrator code), human approvals via external events, per-role RetryPolicy settings, and deployment to Azure Durable Task Scheduler.
Activity for this story
Not enough continuous observations to show a trend yet.