[Experimenting] Add Sentry AI monitoring for OpenAI backend

[jboolean]

Summary

Add Sentry AI monitoring instrumentation around backend OpenAI story-title generation so LLM activity is traceable in Sentry.

Changes

• Updated backend/src/business/ai/AiService.ts to create a child Sentry span around the OpenAI chat completion call.
• Added gen_ai.request span operation metadata with model information (gen_ai.request.model).
• Captured token usage attributes from OpenAI responses (gen_ai.usage.input_tokens, gen_ai.usage.output_tokens) when available.
• Marked the AI span as internal_error on failure and ensured span completion in a finally block.
• Kept prompt/response content capture disabled (no prompt/output payload attributes recorded).

Testing

• Ran formatter: prettier --cache --write backend/src/business/ai/AiService.ts (via Format tool).
• Attempted lint via npm run lint -- src/business/ai/AiService.ts but project script fails due ESLint CLI/config mismatch (--ext invalid with current ESLint).
• Attempted direct lint via npx eslint src/business/ai/AiService.ts but repository is missing flat config (eslint.config.*) for ESLint v9.
• Attempted type check via npm run types; failed because tsc is not available in the current environment.

---

PR by Bits - View session in Datadog

Comment @DataDog to request changes

[datadog-official]

View session in Datadog

Bits Dev status: ✅ Done

CI Auto-fix: Disabled (last PR commit was more than 4 days ago) | Enable

Comment @DataDog to request changes

[datadog-official]

I can only run on private repositories.