Sam Brenner

fix(llmobs): fix span linking for langgraph nodes

3

Fix for `langgraph>=0.3.22`. This includes expanding our testing range for `langgraph`, and it was determined we support as low as `langgraph==0.2.23`. This is a `chore` as the feature this fix...

feat(llmobs): annotate `toolResult` input on bedrock converse spans correctly

4

Makes sure bedrock converse `toolResult` input content blocks are properly annotated on input to the LLM call. We override the `role` field to explicitly be "Tool Result". Before After When...

fix(anthropic): handle malformed streamed responses gracefully

2

## Checklist - [ ] PR author has checked that all the criteria below are met - The PR description includes an overview of the change - The PR description...

Add in Datadog LLM Observability Documentation

5

This PR adds in Datadog LLM Observability documentation for instrumenting stacks using the Google ADK. Datadog has released auto-instrumentation support for this in the `ddtrace` python package within the past...

ci(llmobs): add evaluation metric assertion helper using `node:assert`

3

### What does this PR do? Follow-up on #6709 to re-enable an evaluation metrics test with a new assertion helper ### Motivation Re-enable test, use `node:assert` instead of `mocha` where...

### Tracer Version(s) 5.75.0 ### Node.js Version(s) 22.17.0 ### Bug Report Leaving this as a community notice while I work out a solution, since I haven't seen a report yet....

chore(llmobs): clean up llm observability span tagger

3

### What does this PR do? Cleans up the LLM Observability span tagger for a better (hopefully) devex by 1. Adding typing where possible 2. Making all functions and members...

ci(integration_frameworks): run integration_frameworks scenario from system tests

5

## Description Attempts to run the `INTEGRATIONS_FRAMEWORK` scenario from system tests MLOB-4735 ## Testing CI runs from this PR ## Risks Exposes us to potential mismatch updates from system tests...

feat(pydantic_ai): support `stream_output` in `pydantic_ai>=1.0.0`

3

**Note:** Most LOC are from regenerating riot lockfiles. ## Description Supports `stream_output` on Pydantic AI agent results, which was added in `pydantic_ai>=1.0.0`. To help accomplish testing this, I've bumped the...

chore(llmobs): skip flaky forking test until tests are rewritten

3

## Description I believe the forking tests for the llmobs service tests are causing timeouts in the CI in this repo, and should be subprocess tests. Should be a simple...