Tracer provider defaults to OTEL default provider for evals

[David Elner (delner)]

Trace scorers query spans via BTQL after the task runs. For spans to be queryable, they must first be flushed from the OpenTelemetry tracer provider.

Previously, force_flush was called on eval_context.tracer_provider, which was nil when users didn't explicitly pass tracer_provider: to Eval.run (the common case). The safe navigation operator (&.) silently skipped the flush. In practice, the OTLP exporter's background thread usually shipped spans in time, but this was a race condition — not a guarantee.

Additionally, the runner maintained its own fallback (eval_context.tracer_provider || OpenTelemetry.tracer_provider) to create spans, but this resolved provider was never persisted, so downstream code couldn't rely on it.

Changes

• Context::Factory#build now defaults tracer_provider to OpenTelemetry.tracer_provider when none is provided, so the context always carries a resolved provider.
• The runner's force_flush call uses a respond_to? guard instead of &., since the global provider may be a ProxyTracerProvider that doesn't implement force_flush (e.g. if Braintrust.init was never called).
• Removed the redundant fallback in the runner's initialize since the context now guarantees a provider.

Impact

Users running trace scorers without passing tracer_provider: to Eval.run should no longer need that workaround. The flush will execute against the correct provider automatically.

[Matt Perpick (clutchski)]

i feel like this would be more readable if case_context was shorter

[David Elner (delner)]

Sure. Practically speaking we could abbreviate the variable name to case because its semantically the same.

[David Elner (delner)]

Ahh whoops, forgot case is reserved. So it'd have to be kase (Ruby convention)