[alexcastgonz (pcbeingused333)]

Purpose

With OPENAI_CHAT_HOST=openaicom and no OPENAICOM_CHAT_MODEL, the app fell back to gpt-3.5-turbo. OpenAI shut that model down on 2026-10-23, so the default chat model now fails.

This changes the fallback to gpt-5.4 in src/backend/fastapi_app/dependencies.py, the OPENAICOM_CHAT_MODEL value in infra/main.bicep, .env.sample and the tests, and adds a test for the fallback.

OpenAI's suggested replacement for gpt-3.5-turbo is gpt-5.6-terra, but the chat requests always send temperature (rag_simple.py, rag_advanced.py, default 0.3), and gpt-5.6-terra defaults to medium reasoning effort, which does not accept it. gpt-5.4 defaults to no reasoning, accepts temperature, and is already the default for the Azure host in this repo.

Does this introduce a breaking change?

[ ] Yes
[x] No

Type of change

[x] Bugfix
[ ] Feature
[ ] Code style update (formatting, local variables)
[ ] Refactoring (no functional changes, no api changes)
[ ] Documentation content changes
[ ] Other... Please describe:

Code quality checklist

• The current tests all pass for the file touched (tests/test_dependencies.py, 5 passed; the full suite was not run).
• I added tests that prove my fix is effective or that my feature works
• I ran python -m pytest --cov to verify 100% coverage of added lines (not run; the only non-test change is one string literal in dependencies.py, which the new test exercises)
• I ran python -m ty check to check for type errors (not run)
• I either used the pre-commit hooks or ran ruff manually on my code (ruff check and ruff format --check on the touched files)

This change was prepared with AI assistance (Claude Code).