The change matters less as a new dashboard feature than as an instrumentation boundary: teams can connect model and tool activity to the traces they already use to understand application behavior.
The record begins a distinct analytical branch because no prior New Runtime pattern is supplied that specifically tracks observability coverage and content-mode choices for hosted agent sessions.
Thesis
OpenTelemetry support makes Claude request and agent activity inspectable within a team’s existing observability workflow.
Observed facts
Anthropic documents tracing for Messages API requests, Claude SDK clients, and Managed Agents sessions.
1The documentation says metadata-only is the default content mode and full-content mode includes prompts and responses.
1SDK and API spans can be combined in one trace, while API spans for Claude Platform on AWS are separate.
1
Mechanism
Teams configure SDK tracing, API tracing, or both and send traces to OTLP-compatible observability destinations; content mode determines whether prompts and responses are included.
Why now
As teams operate multi-step agent sessions, debugging a failed or slow request requires visibility beyond a single model response.
Implications
Teams can correlate Claude activity with application traces.
Teams must choose whether prompt and response content enters telemetry.
SDK and API tracing have different setup paths and permission requirements.
Evidence boundary
The public documentation describes available capabilities and setup, not adoption, reliability improvement, or reduced debugging time.
Watch conditions
Whether teams can preserve useful end-to-end traces while maintaining chosen content and access controls.
Whether coverage and trace correlation remain clear across SDK, API, and AWS-hosted requests.