AI API reliability gaps detected

[miguelmanlyx]

Hi 👋

I ran ai-api-check on this repository and noticed a few production-readiness gaps in the AI API integration:

• no timeout protection on AI API calls
• no retry / backoff handling
• no Retry-After header handling
• no request-ID or latency tracing
• no AI API reliability check in CI

These gaps may matter if your provider rate-limits, times out, or has an outage.

How to reproduce

npx ai-api-check

If you ever want provider flexibility, making the base URL configurable via an env var is usually a safe first step — no behaviour change when the env var is unset.

Feel free to close this if these are already handled elsewhere or not relevant to your use case.