Critic Score

A critic score is a 0-to-1 numerical rating assigned to each piece of pipeline-authored content by a separate reviewer model before publication. The score evaluates the draft against a structured rubric: factual accuracy against the underlying fact contracts, voice consistency with the publication's editorial voice, internal coherence, and absence of the "AI tells" the publication's voice constitution explicitly bans.

The score does not gate every surface the same way. The homepage carries a hard floor, currently 0.72, and a homepage draft that scores below it does not publish. It also fails outright if the reviewer cannot confirm the draft's sourcing or finds loaded language in it. Everywhere else the score is advisory: the draft publishes, and a weak score flags it into a queue the analyst works through rather than holding the page empty.

The critic is not a quality oracle. It is a defense against specific failure modes, and drafts that clear it still get read by a person. When it holds back a homepage draft that turns out to be sound, that draft waits in the review queue until the analyst releases it deliberately; the floor itself has no bypass. The critic's value is in catching the bad cases, not in certifying the good ones.