feat(cloudflare,vercel-edge): Add support for Google Gen AI instrumentation

[RulaKhaled]

Adds support for Google GenAI manual instrumentation in @sentry/cloudflare and @sentry/vercel-edge.
To instrument the Google GenAI client, wrap it with Sentry.instrumentGoogleGenAIClient and set recording settings.

import * as Sentry from '@sentry/cloudflare';
import { GoogleGenAI } from '@google/genai';

const genAI = new GoogleGenAI({ apiKey: 'your-api-key' });
const client = Sentry.instrumentGoogleGenAIClient(genAI, { 
  recordInputs: true,
  recordOutputs: true 
});

// use the wrapped client with models api
const model = client.models.generateContent({
  model: 'gemini-1.5-pro',
  contents: [{ role: 'user', parts: [{ text: 'Hello!' }] }]
});

// or use chat functionality
const chat = client.chats.create({ model: 'gemini-1.5-flash' });
const response = await chat.sendMessage({ message: 'Tell me a joke' });

[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	9,026	-	8,714	+4%
GET With Sentry	1,424	16%	1,412	+1%
GET With Sentry (error only)	6,107	68%	6,096	+0%
POST Baseline	1,202	-	1,205	-0%
POST With Sentry	518	43%	537	-4%
POST With Sentry (error only)	1,064	89%	1,042	+2%
MYSQL Baseline	3,344	-	3,297	+1%
MYSQL With Sentry	491	15%	437	+12%
MYSQL With Sentry (error only)	2,708	81%	2,709	-0%

View base workflow run