Is there a way to enable profiling when using the Sentry Lambda layer instead of the SDK?

[micthiesen]

Problem Statement

Currently the docs only show enabling a Sentry extension via arguments to Sentry.init. But if you're using a Lambda layer, then you never call Sentry.init yourself. I don't see a way to enable other integrations using the layer; any ideas? A little surprising since this is the recommended method for setting up Sentry with Lambdas in the docs.

Solution Brainstorm

Currently you can enable tracing via the SENTRY_TRACES_SAMPLE_RATE: "0.5" env var. Not sure if something similar could work here, although this doesn't seems sustainable for arbitrary integrations.