fix: fire response complete on empty responses

[ugur-vaadin]

Description

A turn that produces no visible text — typically a tool-only turn where the model emits tool calls and stops without commentary — is still a successful exchange. But the response-complete lifecycle hooks were skipped: both the application-facing ResponseCompleteListener.onResponseComplete and the controller-facing AIController.onResponseComplete were gated behind a non-empty-text check. Controllers that flush their pending UI state in onResponseComplete (e.g. ChartAIController applying its pending query) never render on these turns; application listeners miss the completion entirely.

This PR moves both hook invocations out of the non-empty-text gate so every successful exchange fires them, with the firing cadence unchanged (still once per user prompt, not per tool call or round-trip). The conversationHistory add stays gated to avoid polluting history with empty assistant entries.

Type of change

• Bugfix
• Feature

Checklist

• I have read the contribution guide: https://vaadin.com/docs/latest/contributing/overview
• I have added a description following the guideline.
• The issue is created in the corresponding repository and I have referenced it.
• I have added tests to ensure my change is effective and works as intended.
• New and existing tests are passing locally with my change.
• I have performed self-review and corrected misspellings.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud