Fix JSON serialization and add corresponding test

[nambili-samuel]

Why are these changes needed?

Related issue number

Checks

• I've included any doc changes needed for https://microsoft.github.io/autogen/. See https://github.com/microsoft/autogen/blob/main/CONTRIBUTING.md to build and test documentation locally.
• I've added tests (if relevant) corresponding to the changes introduced in this PR.
• I've made sure all auto checks have passed.

[nambili-samuel]

Fix JSON serialization and add corresponding test

Description:
This PR addresses a bug in the MCP tool (autogen_ext/tools/mcp/_base.py) where JSON serialization did not set ensure_ascii=False.

Problem:

Non-ASCII characters, including Japanese, were escaped as Unicode codepoints (e.g., "\u65e5\u672c").
Escaped characters reduce readability and can degrade LLM understanding and generation quality in non-English environments.

Solution:

Updated JSON serialization in _base.py to include ensure_ascii=False.
Added a new test (test_mcp_json_serialization.py) to verify that non-ASCII characters (e.g., Japanese text) are serialized correctly without escaping.

Files Changed:

(1) python/packages/autogen-ext/src/autogen_ext/tools/mcp/_base.py
(2) python/packages/autogen-ext/tests/tools/test_mcp_json_serialization.py

Impact:

Fixes JSON serialization for non-ASCII characters. Ensures improved LLM compatibility in multilingual environments.

Testing:

Added unit tests for Japanese characters serialization.
Verified existing functionality is unaffected.