Release v0.2.4

What's Changed

Features

• Distributed Tracing for remote scorers in #33
• Create one span per scorer in evals

Bug Fixes

• Fix an issue where the Braintrust UI would flag score spans as missing a socre

Full Changelog: v0.2.3...v0.2.4

Release v0.2.3

What's Changed

Features

• add support for remote scorers in evals+devserver in #32

Full Changelog: v0.2.2...v0.2.3

Release v0.2.2

What's Changed

Features

• Instrument Langchain4j AI Services in #30

Full Changelog: v0.2.1...v0.2.2

Release v0.2.1

What's Changed

Features

• Initial Support for Remote Evals Devserver #23
  • Dataset fetching
  • Devserver-side task execution
• Instrumentation for Lanchain4j's OpenAI Chat Model #24

Full Changelog: v0.2.0...v0.2.1

Release v0.2.0

Breaking Change Notice:
This release includes a minor breaking change which alters the signature of the Scorer.of(...) factory method

What's Changed

Features

• Support fetching datasets from Braintrust in #20
• Include braintrust.parent in distributed traces in #21
• Link evals to their originating datasets in #22

Full Changelog: v0.1.3...v0.2.0

Release v0.1.3

What's Changed

Features

• Allow configuring custom SSL settings for Braintrust Span Exporter #17
• Support eval dataset tags and metadata #18

Full Changelog: v0.1.2...v0.1.3

Release v0.1.2

What's Changed

Bug Fixes

• log debug instead of info when instrumenting genai client #15

Full Changelog: v0.1.1...v0.1.2

Release v0.1.1

What's Changed

Features

• add time to first token metric for anthropic instrumentation #13
• Instrumentation for Gemini via Google GenAIClient library #14

Full Changelog: v0.1.0...v0.1.1

Release v0.1.0

Breaking Change Notice:
This release includes a minor breaking change affecting users who implement custom task runners or scorers

• To resolve it, update your implementations to include the new Task and Score interface methods
• Please refer to the latest Javadoc for details or open a GitHub issue if you need help.

What's Changed

Features

• add a Builder to BraintrustConfig
• enhance OAI completions instrumentation with additional attribute tagging
• enhance anthropic messages instrumentation with additional attribute tagging
• support Datasets and additional output data in Evals
  • NOTE: remote datasets are not yet implemented; the new interface is designed to support them in a future release

Bug Fixes

• allow passing lambdas to eval task builder without casting
• remove v from maven version prefix by

Full Changelog: v0.0.2...v0.1.0

Release v0.0.2

What's Changed

Features

• support for base64 open telemetry attachments
• Prompt Fetching: https://www.braintrust.dev/docs/platform/functions/prompts#fetching-in-code

Bug Fixes

• improve printing the braintrust URI for self-hosted endpoints with nonstandard ports

Full Changelog: v0.0.1...v0.0.2