ModelGuard
A behavioral contract monitor that runs your MUST-DO, MUST-REFUSE, and FORMAT rules live against an LLM and scores the drift.
ModelGuard lets you define behavioral contracts for your LLM product — rules the model must always follow, must always refuse, and structural format checks — then runs them live against the API to show exactly which contracts break after a silent model update. It ships with a pre-loaded Fable 5 demo so you can paste contracts, run them, and watch a Drift Score reveal what changed. Pure single-file HTML and JavaScript with no backend; bring an OpenAI or Gemini key and everything runs in-browser, with the key kept in memory only.
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