Dusk treats invalid actions as a liability, not a statistic
Visibility is cheap on a blockchain.
Irreversibility is easy to claim.
What is expensive is carrying invalid state forward and pretending it does not matter.
Dusk starts from that assumption.
Instead of tolerating failed or borderline actions as normal network behavior, the protocol treats them as liabilities. Anything that cannot be justified under the rules is filtered out before it is allowed to settle. No revert, no exception, no historical residue.
This is not about being stricter.
It is about being deliberate.
When invalid actions make it into history, they create long-term costs that are rarely measured: monitoring overhead, audit complexity, interpretation risk. The ledger may be final, but the meaning of what happened is not.
Dusk eliminates that ambiguity at the protocol layer.
By enforcing constraints before state exists, the network ensures that settlement already implies compliance. What enters the ledger does not need retroactive explanation. It can stand on its own without context reconstruction.
This is why Dusk does not optimize for visible activity.
Activity that requires cleanup is not a signal of health.
The system is built so that once something becomes state, the question is already answered. Not by governance, not by off-chain review, but by design.
On Dusk, correctness is not enforced after the fact.
It is a prerequisite for existence.
