What it does
An MCP server that connects your AI assistant to OpenTelemetry trace backends such as Jaeger, Tempo and Traceloop. You can ask it to look up traces, find errors, compare model performance and track token usage, with specific support for LLM traces.
Use cases
- 01Find the most expensive LLM calls in a trace backend
- 02Debug errors by asking about failing traces
- 03Compare model performance and token usage