v13.3.0
📦 new-relic-sdkView on GitHub →
✨ 2 features🐛 5 fixes🔧 6 symbols
Summary
This release adds support for AI Monitoring server-side configuration and context propagation for Bedrock invocations via Agent Framework. It also includes several bug fixes, including addressing issues with LangChain, Wagtail, Kombu, and garbage collection sampling.
✨ New Features
- Add support for AI Monitoring server-side configuration. Server-side configuration values for `ai_monitoring.enabled`, `ai_monitoring.record_content.enabled`, and `ai_monitoring.streaming.enabled` now take precedence over local settings.
- Add context propagation for Bedrock invocations through Agent Framework. Trace context is now propagated across thread boundaries created by Microsoft Agent Framework's Bedrock integration, enabling the capture of LLM events for Bedrock invocations made through Agent Framework.
🐛 Bug Fixes
- Fix inconsistent transaction names for apps built on Wagtail, which previously could be the same for different page views.
- Fix duplicate tool entities in the UI for LangChain tools backed by a remote MCP server by gating subcomponent attributes off of LangChain tools when a known remote tool is detected.
- Fix missing User Input field for LangChain and LangGraph MCP tool calls when the input message was not the last in the streamed sequence. The agent now correctly extracts the input message in these cases.
- Reduce race condition risk in the garbage collection data sampler by only calling `gc.get_objects()` when `top_object_count_limit` is configured above 0, avoiding unnecessary heap iteration.
- Remove byte encoding from outbound Kombu distributed tracing headers, which previously broke transports that serialize payloads with the standard library `json` module.