Does the groundedness classifier only work on agents that use RAG?

No. It scores any LLM call against whatever context that call actually received, tool outputs and prior conversation turns included, not only documents pulled by a retrieval step.

Retrieval is the most obvious source of context to check an answer against, but it isn’t the only one. An agent that calls a tool, gets a result back, and then states something the result doesn’t support has the same failure shape as a bad RAG citation, just with a different source of context. A trip-planning agent asserting a flight time no tool ever returned is as ungrounded as a support bot misquoting a document.

What the classifier looks at is only what was actually passed into that specific call. An agent with no retrieval step and no tools, answering from a system prompt and the conversation so far, still has a context to be checked against, and can still be scored on whether its answer stays inside it.

keep reading

More on this.

Can a factually correct answer still get flagged as ungrounded?Yes. Groundedness checks whether an answer is supported by its context, not whether the answer is true, and those are different questions.answer →Does groundedness check every LLM call, or just the final answer?Every LLM call. It scores each call against its own context as the call happens, not just the last message a trace sends back to the user.answer →Does the groundedness classifier check whether the context itself is correct?No. It checks whether the answer matches the context given, not whether that context is accurate, so a wrong or stale source can still score grounded.answer →Everything on tessary groundednessThe definition, and every question answered under tessary groundedness.concept →

Send us the traces you already emit.

book a demo