Every release begins here, as raw matter — quote tape, volume prints, sector ETF flows, supply-chain releases. None of it is editorial yet. The gate decides only whether it belongs in a registered bucket at all.

Raw market data enters the bucket-classifier and must match a registered bucket signature or be refused. No bucket is invented at runtime; no signal is admitted without an intake print of record.

Classified buckets hold one of three wave states — active, cooling, re-accelerating.

Assignment is mechanical from intake metrics, not editorial. Transitions between wave states are themselves attested events, dual-signed and entered to the ledger before any downstream gate is entered.

A grade between one and five — not editorial preference.

Active buckets earn a conviction grade between one and five. The grade certifies the depth of the citation chain backing the signal, and falls as fast as it rises if cited prints are superseded.

The binding sentence of the directive — every published thesis cites at least one tape print.

The gate witnesses every passage; it does not infer, predict, or estimate. A thesis without a tape-citation set on file is refused publication, regardless of conviction grade.

A thesis that has passed the prior four gates is published — as a record, not as a recommendation.

The release is signed by its witnessing personas and entered to the Signal Map under the timestamp of its sealing witness. Vyzora Lens publishes editorial research, not investment advice.