new·The score now tells you which way it movedA brain's exam only ever grows: its own material writes questions, and so does every question a real caller asked and did not get answered. The score is a percentage over that growing set, so a brain that learned more could post a smaller number — and this week three did. One of them answered two MORE questions than the week before and showed eighteen points less. Printed as a single percentage, that reads as decline to a reader and as punishment to anyone who contributes material.all news →
mozg.beta
Sign in

AI SDK · Providers · all subjects

langwatch/capabilities

5 notes, read out of this brain and free to use. Each one was extracted from a source and is re-checked against its exam.

LangWatch trace concept

Each message triggering your LLM pipeline as a whole is captured with a Trace. A Trace contains multiple Spans, which are the steps inside your pipeline.

LangWatch span types

A span can be an LLM call, a database query for a RAG retrieval, or a simple function transformation. Different types of spans capture different parameters and can be nested to capture the pipeline structure.

LangWatch manual integration with getTrace

For manual integration without OpenTelemetry, initialize LangWatch and use getTrace method with metadata containing threadId and userId.

LangWatch startLLMSpan method

Start an LLM span within a trace using trace.startLLMSpan() with name, model, and input object containing type 'chat_messages' and value array of messages.

LangWatch span.end() method

End a span using span.end() with output object containing type 'chat_messages' and value array, plus metrics object with promptTokens and completionTokens for cost calculation.

Give your agent this brain