Does the groundedness classifier check whether the context itself is correct?

No. It checks whether the answer is supported by the documents the agent was given, not whether those documents are accurate, so an answer built faithfully on a wrong or stale source still scores grounded.

Groundedness answers one narrow question: did the agent stay inside what it was handed, or did it add something the documents never said. If retrieval returns last quarter’s pricing page and the agent repeats that price exactly, the answer is grounded, and it’s the retrieval that was wrong. That’s a different failure, a stale index, upstream of anything this classifier is built to catch.

The model has no world knowledge to compare against, only the text it’s passed. Judging whether each document is current and correct would need a source of truth for every one of them, which is a different problem from noticing when an agent says something its own documents don’t support.

keep reading

More on this.

Does the groundedness classifier only work on agents that use RAG?No. It checks RAG answers against their retrieved documents, and summarize or extract calls against their prompt. RAG answers built only from tool output aren't checked.answer →How long before Tessary's groundedness classifier starts flagging answers?It starts once its reference holds 100 traces, down from 200; it already kept learning until 1,000 before this change, and still does.answer →What is Tessary's groundedness classifier?A built-in classifier that checks your agent's answers against the documents it retrieved, or the prompt when it retrieved none, and flags the sentences that source doesn't support.answer →Everything on tessary groundednessThe definition, and every question answered under tessary groundedness.concept →

Two ways to run Tessary.

Tessary is an open-source agent reliability platform. Cloud and self-hosted run the same workflow on the OpenTelemetry traces your agent already emits.

Tessary Cloud

We host it for you. Send your first trace with nothing to deploy and no model key.

what's includedper organization
traces
10,000 per calendar month
stored trace data
1 GB
retention
30 days
model credit
$10, one-time, for triage and root-cause analysis
credit card
not required
start for free

Self-hosted Tessary

Run the open-source code on your own infrastructure with one command. Add your own model key for triage and root-cause analysis.

Self-host Tessary for me by following https://github.com/tessaryai/tessary/blob/main/setup.md

docker compose -f oci://docker.io/tessaryai/tessary:compose up -d -y