[kaxil]

Teams with a Pydantic AI agent of their own had no guide: the docs covered AgentOperator and assumed the operator runs the agent. The new Pydantic AI page shows the model from PydanticAIHook.get_hook(...).get_conn(), the same vendor handling AgentOperator uses, and the provider's toolsets passed straight to the agent. It lists what AgentOperator adds on top: durable replay, human review and tool-call approval, masking for the Agent Skills toolset and your own toolsets, templated connection IDs, and tool logging with the task's identity on spans. For spans, it points to the hook's create_agent, which instruments the agent the way AgentOperator does, with prompt text off by default.

The sandbox guide now shows a Strands agent owning its sandbox's life with a with block.

This is the top of the native agent framework stack; nothing else depends on it.

The new guide as rendered:

---

• Read the Pull Request Guidelines for more information. Note: commit author/co-author name and email in commits become permanently public when merged.
• For fundamental code changes, an Airflow Improvement Proposal (AIP) is needed.
• When adding dependency, check compliance with the ASF 3rd Party License Policy.
• For significant user-facing changes create newsfragment: {pr_number}.significant.rst, in airflow-core/newsfragments. You can add this file in a follow-up commit after the PR is created so you know the PR number.