Agent reliability

Instrumentation

Instrumentation is the code that records what an agent does while it runs and emits that record as telemetry. For agents, the standard is OpenTelemetry with the GenAI semantic conventions. A framework that already emits OTel spans exports them over OTLP, via HTTP or gRPC, to any endpoint that accepts the protocol. Other dialects, such as OpenInference and Traceloop, describe the same events with different attribute names.

Agent telemetry has a three-level structure that follows the shape of agent execution. A session groups everything that happened across one conversation. A trace covers one turn: a user message, the agent's full response, and everything between. A span covers one step within the turn, an LLM call or a tool call. Spans nest within traces, and traces group into sessions.

Instrumentation is also the point where content leaves the application. It decides what gets emitted: which prompts, outputs, and tool arguments appear in the record, and in what form. The telemetry downstream is whatever instrumentation chose to emit, so content decisions, including redaction, happen here.

6 questions

Answered, plainly.

Can the same OpenTelemetry traces go to more than one tool?Yes. An OpenTelemetry Collector can receive one stream of spans and fan it out to several exporters, each pointed at a different backend.answer →Do agent traces have to use the OpenTelemetry GenAI attribute format?No. OpenInference and Traceloop describe the same LLM and tool calls with different attribute names; gen_ai.* is the standard, not the only option.answer →Does emitting OpenTelemetry traces slow down an agent's response?No, not when it's set up correctly. Span export runs asynchronously in a batch processor, off the thread that produces the agent's answer.answer →Should agent traces export over HTTP or gRPC?Either works under OTLP: gRPC defaults to port 4317, HTTP with protobuf to port 4318. Teams often pick gRPC for overhead, HTTP for proxies and firewalls.answer →What is agent instrumentation?The code that records what an agent does while it runs and emits that record as telemetry, usually OpenTelemetry spans sent over OTLP.answer →Who decides what content leaves an agent's traces?Instrumentation does. It chooses what a span carries, prompts, outputs, tool arguments, before export, so redaction has to happen at or before that point.answer →

Send us the traces you already emit.

book a demo