Structured findings, not objective grades.

What it does

Evaluation is the background process that produces or refreshes observations. Structured fields can describe dimensions such as awareness, tension, voice, trajectory, or a project-specific metric, while evidence and confidence keep the finding inspectable.

Why it matters

Comparable values help an author notice movement and contrast across scenes. They do not establish a universal theory of quality or tell the author what must be fixed.

• Canonical observations are stored under ai.observations.
• Metric notes supply the author’s interpretive meaning.
• Configuration controls the eligible targets and bounded context.
• Custom dimensions can express the questions a particular story needs.
• Saved observations become durable material for ObservaStory workspaces and lenses.