Every rule in the mapping registry, read only. Editing happens where an editor exists.
Three different facts, and they are worth keeping apart.
Stored means the row is in the registry. Active means its status is active and it carries a published revision.
Read by an engine means some code actually loads it and acts on it. A rule can be stored and active and still be read by nothing,
which is the case for most of what is here.
MWE policy controls are incomplete.
The term-lookup prompt has a registry-backed editor. Correction policy has been registered, but its revised
implementation and operator controls are not yet complete. Eligibility, discovery and ranking remain implemented in
code, in candidates() and the SCORING policy in
assets/citations-render.js and in the pack build, without workbench policy controls. A developer can
read and change that code; an operator cannot direct it from here, and that is the gap.
Every automated MWE decision must have an inspectable policy, human controls, an explanation of its effect, and a way
to correct the result. Policy changes must be previewable, versioned, tested and reversible. This requirement remains
unfinished until operators can use those controls.
Editing the term-lookup prompt changes adjudication after a phrase is proposed. It does not change which phrases
discovery finds or how ranking orders them. Those controls remain required work under DR-094.
The three tags on each rule below are three different facts. A registry row on its own satisfies none of the others:
status says whether the rule is active, read by an engine says whether code actually consults it, and
operator-editable says whether there is a control in this workbench for changing it.