Does a passing unit test rule out a regression?

No. A unit test checks that your code did the right thing mechanically, such as calling a tool with a schema-valid payload or handling an exception without crashing, and it can pass every time while the agent’s actual behavior gets worse. Change a tool’s response format and a well-written test confirms your code still parses it; it says nothing about whether the model, reading that new format, still draws the same conclusions from it. The two questions are different: did the code execute the right path, and did the model behave well once it got there. A dependency bump or a config edit can leave every unit test green while quietly shifting what the model does with the result, because none of those tests were ever asserting anything about model behavior to begin with. Catching that needs a check on the agent’s actual output, not on the code path that produced it.

keep reading

More on this.

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

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