EXPERIMENTAL
HUMAN / SEPARATE PROCESSPREDICTED
A declared public rule/model may generate a PREDICTED candidate, but the original experimental representation remains unchanged.
LABS METHOD
Experiments do not promote themselves.
Labs outputs never self-promote into canonical relations or VERIFIED claims.
PREDICTED
HUMAN / SEPARATE PROCESSREVIEWED_EDITORIAL
A human editorial review must inspect the candidate basis, canonical profiles and relevant context before any public relation change is considered.
REVIEWED_EDITORIAL
HUMAN / SEPARATE PROCESSVERIFIED
Editorial review cannot create VERIFIED evidence. A separate evidence-backed claim with appropriate external sources and scope is required.
No automatic graph mutation
Adding or changing an explicit public relation requires a deliberate source-data/editorial change outside the Labs candidate generator.
Verification is claim-specific
VERIFIED status applies only to claim records linked to the public source registry; candidate links cannot inherit it.
Methods stay inspectable
The public Labs export should expose method and boundary information sufficient to understand how an experimental record was generated.
A compelling Labs result can remain experimental indefinitely. There is no automatic threshold that converts it into truth.