cleandataengineering.aiAgent Harness, drawn simply
You are here: End to end › Checks › 3 · After the model › no_superseded_state

Agent Harness Tutorial · End to end · Zoom level 3

How we take care of it: no_superseded_state

no_superseded_state catches an old status after a newer one.

Open: 11before loopopen ›Open: 22before modelopen ›Open: 33after modelopen ›Open: 44before toolopen ›Open: 55the toolopen ›Open: 66after toolopen ›Open: 77after loopopen ›Open: 88before replyopen ›It readsthe reply, and the newest statusread for each recordno_superseded_staterule · agentpassCarries onfailIf it failssafe replyWHERE IN THE CODE: F policy rules · evaluators.yaml reply
The strip at the top shows every check point where this check runs, in dark.

How we took care of it.

  1. If the reply states a status, it must be the newest one read, not an older one from earlier in the turn.
It readsthe reply, and the newest status read for each record
Kindrule
Provided byagent
If it failssafe reply
Where in the codeF policy rules · evaluators.yaml reply
Runs at3 · After the model
Real examplePairs with the re-read of earlier rows in context assembly.

More on this: What goes in.