[Feat]: Samples allowed for Custom Model URLs

[jland-redhat]

Is your feature request related to a problem? Please describe.

I am setting up an Active-to-Active (A2A) demo environment. I have an existing vLLM deployment with a model accessible within this environment.

The current A2A samples in this repository appear to be primarily designed for use with Gemini models. While I understand this may not be the central focus of the samples, it would be nice to be able to simplify it for users like myself who wish to quickly show a demo with the services integrated to their own enterprise models or alternative large language models (LLMs) other than Gemini.

Describe the solution you'd like

I think it would be nice and relatively easy adding a small amount of additional logic to the A2A samples to allow for more flexible model endpoint configuration. Specifically, it would be beneficial to support the inclusion of OPENAI_URL and OPENAI_TOKEN environment variables. If these variables are present and a Gemini API key is not found, the samples should ideally default to using the specified OpenAI-compatible endpoint. This would significantly ease the deployment and testing of A2A scenarios with custom or self-hosted LLMs.

Describe alternatives you've considered

No response

Additional context

No response

Code of Conduct

• I agree to follow this project's Code of Conduct

[flystar32]

As I've mentioned on #330 (comment)
it's works for me,wish it can help you