Sentry SDK 20.10.0 change: breadcrumbs omitted in containerized CI lacking HTTP request context

The goal is to retain visible breadcrumbs for errors that occur during repeatable, containerized CI builds. In these environments the HTTP request context that Sentry normally uses to auto‑capture breadcrumbs is frequently absent, and SDK version 20.10.0 altered the default behavior of the attach_stacktrace option to omit stack traces for certain SDKs unless explicitly opted in.

Because breadcrumbs are only generated from HTTP requests by default, the combination of missing request context and the new attach_stacktrace default can lead to silent loss of diagnostic data. Teams must decide how to restore breadcrumb visibility without introducing manual instrumentation that could drift across pipelines.

• Should attach_stacktrace be enabled globally for all SDKs used in CI to ensure stack traces are retained?
• Is manually adding breadcrumbs via Sentry.addBreadcrumb the recommended approach for capturing CI‑specific events?
• Does manually setting the request context (e.g., with Sentry.configureScope) restore automatic breadcrumb capture in containerized environments?