fix: resolve message duplication in LangChain integration

[OhYee]

• Convert AgentRunConverter.to_agui_events() from static to instance method calls in tests
• Add comprehensive integration tests for LangChain/LangGraph invoke methods
• Add conversion tests for Python 3.10 and 3.12 compatibility
• Ensure proper state management to prevent duplicate message output in streaming scenarios

Change-Id: I113237eeb2cb73bf0e7b00e9e55882f2f32e6c10

Thank you for creating a pull request to contribute to Serverless Devs agentrun-sdk-python code! Before you open the request please answer the following questions to help it be more easily integrated. Please check the boxes "[ ]" with "[x]" when done too.
Please select one of the PR types below to complete

---

Fix bugs

Bug detail

The specific manifestation of the bug or the associated issue.

Pull request tasks

• Add test cases for the changes
• Passed the CI test

---

Update docs

Reason for update

Why do you need to update your documentation?

Pull request tasks

• Update Chinese documentation
• Update English documentation

---

Add contributor

Contributed content

• Code
• Document

Content detail

if content_type == 'code' || content_type == 'document':
    please tell us `PR url`，like: https://github.com/Serverless-Devs/agentrun-sdk-python/pull/1
else:
    please describe your contribution in detail

---

Others

Reason for update

Why do you need to update your documentation?

[Copilot]

Pull request overview

This PR makes the LangGraph/LangChain → AG-UI conversion stateful per converter instance to avoid duplicate message emission across different event formats, and adds comprehensive integration and compatibility tests (including Python 3.10 vs 3.12 event shapes and multiple LangChain invoke patterns). It also updates AG-UI and OpenAI-protocol expectations in tests to match the new, de-duplicated streaming behavior.

Changes:

• Convert AgentRunConverter.to_agui_events from a static-style API to an instance method, introduce has_on_chat_model_stream to gate on_chain_stream handling, and adjust internal conversion helpers accordingly.
• Update existing integration tests to construct a converter instance and expand coverage with new LangChain/LangGraph invoke-mode tests and Python 3.10/3.12 event-shape fixtures.
• Align e2e expectations (AG-UI and OpenAI streaming/non-streaming) with the new tool-call and text-message event ordering to ensure no duplicated tool or text outputs.

Reviewed changes

Copilot reviewed 9 out of 9 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
agentrun/integration/langgraph/agent_converter.py	Makes the converter fully instance-based, adds has_on_chat_model_stream to suppress duplicate on_chain_stream handling, and wires all internal converters through self.
tests/unittests/integration/test_langgraph_to_agent_event.py	Switches tests to use AgentRunConverter() instances instead of the previous static-style to_agui_events calls.
tests/unittests/integration/test_langgraph_events.py	Same instance-based adjustment for LangGraph-focused converter tests.
tests/unittests/integration/test_langchain_convert.py	Updates existing LangChain conversion tests to the instance API and ensures streaming behavior assertions use a per-test converter instance.
tests/unittests/integration/test_agent_converter.py	Mirrors the instance-API refactor across the broader converter test suite.
tests/unittests/integration/langchain/test_convert.py	New unit tests that feed recorded 3.10 and 3.12 LangGraph/LangChain event sequences through AgentRunConverter.convert and assert de-duplicated final text.
tests/unittests/integration/langchain/test_agent_invoke_methods.py	New integration tests that spin up a mock OpenAI-compatible backend and verify AG-UI/OpenAI outputs across astream_events, astream, stream, invoke, and ainvoke for both plain text and tool-call flows.
tests/unittests/integration/conftest.py	Adjusts the shared convert_and_collect helper to use a converter instance per call, consistent with the new API.
tests/e2e/integration/langchain/test_agent_invoke_methods.py	Updates end-to-end AG-UI and OpenAI expectation fixtures to the new tool-call/text-message ordering and adds Python-version–specific conversion checks similar to the unit tests.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

This section header still labels the following helpers as "静态方法" and the examples further down the file still show calling AgentRunConverter.to_agui_events(event) on the class, but to_agui_events has been converted to an instance method. To avoid confusing users, please update the comment and examples to describe and demonstrate instance-based usage (for example, via a converter instance) instead of static calls.