Lee-W commented on code in PR #74026: URL: https://github.com/apache/airflow/pull/74026#discussion_r4258946294
########## providers/common/ai/docs/frameworks/claude_agent_sdk.rst: ########## @@ -0,0 +1,195 @@ + .. Licensed to the Apache Software Foundation (ASF) under one + or more contributor license agreements. See the NOTICE file + distributed with this work for additional information + regarding copyright ownership. The ASF licenses this file + to you under the Apache License, Version 2.0 (the + "License"); you may not use this file except in compliance + with the License. You may obtain a copy of the License at + + .. http://www.apache.org/licenses/LICENSE-2.0 + + .. Unless required by applicable law or agreed to in writing, + software distributed under the License is distributed on an + "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY + KIND, either express or implied. See the License for the + specific language governing permissions and limitations + under the License. + +.. _howto/frameworks:claude_agent_sdk: + +Claude Agent SDK +================= + +If your agent is a loop on the `Claude Agent SDK +<https://platform.claude.com/docs/en/api/agent-sdk/overview>`__'s ``query()``, which runs +the Claude Code CLI as a subprocess, you can run it in an Airflow task as it is and give +it the toolsets this provider ships. +:class:`~airflow.providers.common.ai.tools.claude_agent_sdk.AirflowTools` serves +``SQLToolset``, ``HookToolset`` and most other toolsets to the CLI through an in-process +MCP server. If you have no loop yet, ``AgentOperator`` runs one for you, with durable +replay and human review; see :doc:`index`. + +.. note:: + + ``AirflowTools`` and the framework-neutral tool interface it builds on are + experimental. They may change in a minor release of this provider. + See :ref:`howto/stability`. + +Run a loop in a task +--------------------- + +This task answers a question about a database. The client's API key comes from an +Airflow connection: + +.. exampleinclude:: /../../ai/src/airflow/providers/common/ai/example_dags/example_claude_agent_sdk.py + :language: python + :start-after: [START example_claude_agent_sdk] + :end-before: [END example_claude_agent_sdk] + +The model called one tool, ``lookup``. Here is the task log's tool-call group, Review Comment: Replaced both blocks with output from a real run of the example against a throwaway SQLite database: the log block is the task log of the agent calling `list_tables`, `get_schema` and `query`, and the `is_error` block is what the model read when asked to write. <img width="1104" height="383" alt="image" src="https://github.com/user-attachments/assets/11ce51fa-165b-428b-847d-57108543f4c9"; /> <img width="1040" height="110" alt="image" src="https://github.com/user-attachments/assets/3f767cb0-cc4a-46de-bd04-3cb6fd58f394"; /> -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
