Allow using Gemini 2.5 Flash for code generations

What does this MR do and why?

Introduce a new feature flag use_gemini_2_5_flash_in_code_generation to control whether Gemini 2.5 Flash should be used for code generation.

References

Related to #549161 (closed)

How to set up and validate locally

1. Checkout this branch in GitLab project
2. Checkout the branch from gitlab-org/modelops/applied-ml/code-suggestions/ai-assist!2806 (merged) in the AI GW project
3. In the GitLab project, run bundle exec rails r "Feature.enable(:use_gemini_2_5_flash_in_code_generation)"
4. Change the GitLab account in your workspace to your local instance (e.g. https://docs.gitlab.com/editor_extensions/visual_studio_code/#switch-gitlab-accounts-in-vs-code)
5. In your editor, trigger a code generation by adding a comment with instructions to GitLab Duo
6. You should see the log below in you AI GW logs:

AI GW Logs

2025-06-20_14:44:31.21644 gitlab-ai-gateway       : 2025-06-20 15:44:31 [info     ] Returning prompt from the registry correlation_id=01JY6WRR4F6XW4BKSNDJV48QWF model_identifier=None prompt_id=code_suggestions/generations/base prompt_name='Gemini 2.5 Flash Code Generations Agent' prompt_version=1.2.0-dev

MR acceptance checklist

Evaluate this MR against the MR acceptance checklist. It helps you analyze changes to reduce risks in quality, performance, reliability, security, and maintainability.