fix(smolagents): adds reasoning_content

[bruffridge]

resolves: #1641
CC: @albertvillanova , @nate-mar

If reasoning_content is present (it should be if a reasoning LiteLLM model is called), add it to the output message.

[github-actions]

CLA Assistant Lite bot All contributors have signed the CLA ✍️ ✅

[bruffridge]

I have read the CLA Document and I hereby sign the CLA

[axiomofjoy]

@bruffridge Can you add a code snippet to the examples directory I can run to verify that this appears as expected in the Phoenix UI? A unit test would also be helpful, I can help there if needed.

[bruffridge]

I added an example.

[axiomofjoy]

I added an example. If you have trouble connecting to anthropic, I can add a different thinking model such as o3 to the example file.

Thanks @bruffridge! Will take a look.

[bruffridge]

I added a unit test. I didn't run it though, so you should check to see if I missed anything.