Clean combines verification evidence, policy, durable bulk operations, API authority and immutable receipts into one inspectable trust system.
Clean keeps observation, interpretation and policy distinct so operators can see what happened instead of receiving an unexplained green or red badge.
Syntax, Unicode/IDN normalization, domain and TLD checks, MX/null-MX/implicit-MX handling, role/disposable/free-provider signals, typo advisories and bounded SMTP evidence.
Durable list cleaning that preserves source context and survives interruption instead of treating uploads as disposable request work.
A versioned REST boundary for putting the same verification authority inside forms, products and data pipelines.
Every accepted decision remains inspectable after the moment it was made.
CoreID is the identity authority. Clean projects organization membership and rechecks sensitive permissions at invocation.
Long-running verification work stays off request processes and is designed to recover cleanly.
Clean uses a small, stable vocabulary so humans and systems can act consistently without flattening uncertainty into a false binary.
The selected policy sees sufficient positive evidence and no suppressing condition.
The address is not automatically suppressed, but one or more signals deserve review.
Observed evidence supports suppression under the selected policy.
Clean refuses to invent an answer when the available evidence does not justify one.
These are design differences in how Clean represents evidence—not unsupported claims about another vendor’s implementation.
Start with one address, then move the same decision model into bulk operations and your application.