[Feature][gpt-oss] Python Tool Test Enhancement

[heheda12345]

🚀 The feature, motivation and pitch

Python tool is disabled in CI now because it needs to run in docker but we can't do it as CI is already inside a docker.

vllm/tests/entrypoints/openai/test_response_api_with_harmony.py

Line 343 in 36c260d

@pytest.mark.skip(reason="Code interpreter tool is not available in CI yet.")

vllm/tests/entrypoints/openai/test_response_api_with_harmony.py

Line 282 in 36c260d

# "What is 13 * 24? Use python to calculate the result.",

Please also find a way to make sure the python tool is called and gives reasonable output.

To enable the test, you can use uv to execute the python tool with PYTHON_EXECUTION_BACKEND=dangerously_use_uv (https://github.com/vllm-project/recipes/blob/46a7a1c57ec04f6118e2fdc47e78f0b858f3180e/OpenAI/GPT-OSS.md?plain=1#L154C323-L154C365)

Alternatives

No response

Additional context

No response

Before submitting a new issue...

• Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.

[lengrongfu]

Use docker in docker can resolve this issue, what do you think. @heheda12345

[wuhang2014]

I think I can do this @heheda12345

[wuhang2014]

Use docker in docker can resolve this issue, what do you think. @heheda12345

The outter docker needs to be privileged, I'm not sure if it will work in a CI environment. Maybe podman in docker is a potential solution....

[heheda12345]

@lengrongfu I feel uv should be easier to setup than docker in docker.