Agent Failure Lab
Run three local failure scenarios and compare the same state machine with and without a guard.
A local synthetic benchmark covering context decay, duplicate effects and budget overruns. Inspect traces, answer a check question, export results and download the code and tests. The 24 paired fixtures demonstrate guards in fixed scenarios, not model accuracy, real incidents or a savings forecast. No API, input collection or automatic storage.
The build log
Follow the work.
New essays and prototypes, in your feed reader. Email subscriptions are paused while I connect a working delivery service. No email is collected here.
Follow via RSS