See exactly what leaves.
Anonymisation Preview shows precisely what a prompt looks like after Redact has processed it, so you can confirm what is protected before anything reaches a model.
Paste or send a prompt and see the anonymised version side by side, so it is obvious which fields were caught and which were left as is.
Preview makes it safe to adjust charters, because you can check the effect on real prompts rather than guessing at what a change will do.
The point of preview is confidence. You see what leaves before it leaves, so protection is something you can verify rather than assume.
Related capabilities on the same authorisation layer. Every one is enforced before an action runs and sealed to the Provenance Vault.
Put one workflow behind Xybern and watch every agent action authorised, and sealed to the vault.
Xybern