Improve error handling for the generate commit message action

What does this MR do and why?

Improve error handling on the Generate commit message feature so it doesn't leave a hanging modal when the API call to the AI Gateway fails (in the case of the related issue, due to a missing prompt).

This improves UX as it shows a nicer error message rather than leaving the user with no information about the progress of their request.

References

Related to #578687

Screenshots or screen recordings

How to set up and validate locally

Temporarily apply the diff below and follow the steps shown in the gif above.

diff --git a/ee/lib/gitlab/llm/ai_gateway/completions/generate_commit_message.rb b/ee/lib/gitlab/llm/ai_gateway/completions/generate_commit_message.rb
index 8789158fdc48..29ef946c0055 100644
--- a/ee/lib/gitlab/llm/ai_gateway/completions/generate_commit_message.rb
+++ b/ee/lib/gitlab/llm/ai_gateway/completions/generate_commit_message.rb
@@ -16,7 +16,7 @@ def inputs
 
           override :prompt_version
           def prompt_version
-            ::Ai::AmazonQ.enabled? ? 'amazon_q/1.0.0' : '1.2.0'
+            ::Ai::AmazonQ.enabled? ? 'amazon_q/1.0.0' : '1.3.0'
           end
 
           override :root_namespace

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.