refactor: Updated @google/genai instrumentation to subscribe to events emitted

[amychisholm03]

This PR resolves #3441

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 97.68%. Comparing base (3ee4d38) to head (f2bfb17).
⚠️ Report is 5 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3467      +/-   ##
==========================================
- Coverage   97.80%   97.68%   -0.12%     
==========================================
  Files         428      431       +3     
  Lines       56592    56773     +181     
  Branches        1        1              
==========================================
+ Hits        55351    55461     +110     
- Misses       1241     1312      +71     

Flag	Coverage Δ
integration-tests-cjs-20.x	73.60% <49.56%> (-0.37%)	⬇️
integration-tests-cjs-22.x	73.64% <49.56%> (-0.37%)	⬇️
integration-tests-cjs-24.x	74.37% <49.56%> (-0.38%)	⬇️
integration-tests-esm-20.x	52.51% <49.56%> (-0.12%)	⬇️
integration-tests-esm-22.x	52.56% <49.56%> (-0.12%)	⬇️
integration-tests-esm-24.x	53.73% <49.56%> (-0.14%)	⬇️
unit-tests-20.x	88.64% <49.56%> (-0.07%)	⬇️
unit-tests-22.x	88.67% <49.56%> (-0.07%)	⬇️
unit-tests-24.x	88.68% <49.56%> (-0.07%)	⬇️
versioned-tests-20.x	80.99% <100.00%> (-0.24%)	⬇️
versioned-tests-22.x	81.00% <100.00%> (-0.24%)	⬇️
versioned-tests-24.x	80.93% <100.00%> (-0.24%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[bizob2828]

Nice work, I had a few comments around some things. I think we can also file a ticket to abstract LLM based instrumentation to have a base LLM class to remove this copy pasta around recording messages, errors, etc but for a different story.

[amychisholm03]

agent.config -> config will be addressed in #3487

[bizob2828]

We may have to re-check the spec for ai monitoring. Right now when the subscriber is registered it checks if ai_monitoring.enabled is set to true. If not, it will not subscribe. Any subsequent checks will never be false. This may be broken in openai as well. You should use this.config to check config, not this.agent.config. Yes they are the same thing but we're trying to create consistency.

[bizob2828]

I'm going to have a lot of the same comments:

• remove optional chaining for checking configuration
• use this.enabled to check if ai monitoring is enabled
• add a this.streamingEnabled and use that to check if streaming is enabled
• We should only be checking if ai monitoring is enabled before creating LLM events, the segment should still be created

[bizob2828]

A few comments inline but also this instrumentation has created some uncovered paths that need tests:

• streaming is enabled but ai monitoring is not enabled
• streaming and ai monitoring are enabled but no active transaction
• embeddings when ai monitoring is disabled
• embeddings when ai monitoring is enabled but not in active transaction

[bizob2828]

LGTM, good work on this one