How to use, and what is real: load the latest transcript from an exhibit or paste JSON manually, then compare transcript semantics and verifier outcomes side by side. For Fiat-Shamir transcripts this page does not take the recorded challenge on trust — it re-derives it with a real SHA-256 digest over the transcript's own (R, y, message) and reports match or mismatch. The remaining rows are field-level comparison of the JSON, not verification: this is a teaching inspector for proof traces, not a proof checker.