Calibration and Transparency
The network's shared public record, and what CUSP publishes so its decisions can be checked against outcomes.
This is the network's shared public record and the calibration ledger for CUSP Risk. Every decision the network makes is published, and every published number can be checked against what actually happened. That is how the Outcome Risk Object earns the trust an outside lender puts in it, not through reporting bolted on after the fact.
What is public
- Score history: every value CUSP Risk assigned, next to the outcome that realized, so calibration is measured, not claimed.
- Diagnostics: how healthy the engine's measurements have been over time.
- Parameter changes, with reasons: every change recorded with its rationale, never applied silently.
- The liquidation log: every auction, cleared or not.
Marks are also published through oracle networks, so external applications can read the same numbers the protocol acts on.
Why honesty is the default
At launch, credit is bounded by first-loss capital, so senior protection is arithmetic. Past that bound, CUSP can only scale as fast as its calibration history justifies, which makes the published record the real constraint on how big CUSP can get. Growing means publishing a clean record, because the record is what permits the growth. The formal scaling posture is in the CUSP whitepaper.