fix(node): Add origin for OpenAI spans & test auto instrumentation

[mydea]

We used to mock the OpenAI client and manually instrument this. This adds a test for actual auto-instrumentation of OpenAI, and adjusts the origin to be fixed.

[github-actions]

node-overhead report 🧳

Note: This is a synthetic benchmark with a minimal express app and does not necessarily reflect the real-world performance impact in an application.

Scenario	Requests/s	% of Baseline	Prev. Requests/s	Change %
GET Baseline	8,854	-	8,872	-0%
GET With Sentry	1,284	15%	1,225	+5%
GET With Sentry (error only)	5,892	67%	5,982	-2%
POST Baseline	1,181	-	1,174	+1%
POST With Sentry	464	39%	518	-10%
POST With Sentry (error only)	1,043	88%	1,046	-0%
MYSQL Baseline	3,307	-	3,241	+2%
MYSQL With Sentry	440	13%	425	+4%
MYSQL With Sentry (error only)	2,674	81%	2,655	+1%

View base workflow run