Fluid checklists can be scored by supplying situation notes from an external knowledge system, assessing each note against the descriptions of a Fluid checklist, and then recording the result through the checklist interface or its equivalent database operations.
For each note, assess every checklist item using this rule:
- Mark complete when the note clearly demonstrates the behavior described by the item.
- Leave both complete and ignored unticked when the note clearly demonstrates the opposite behavior. This records a relevant but unsuccessful observation.
- Mark ignored when the item is not relevant to the note or the evidence is ambiguous. Do not infer failure merely from missing evidence.
Process notes one at a time. Before each note, clear the temporary checkbox
state from the previous note, apply the complete and ignored selections, and
leave relevant opposite-behavior items unticked. Then use the checklist's
Clear all action. For scored checklists, this increments assessed for
each non-ignored item and increments score for each completed item before
clearing the temporary selections.
Do not alter the source notes while scoring. Preserve the distinction between an ignored observation and an incomplete observation: ignored items do not contribute to the denominator, while incomplete items do. If a source note refers to multiple people, apply the assessment only to the person or subject specified by the task.
When using direct database access, update the instance-specific checklist rows
and reproduce the application order: set temporary flags, assess all
non-ignored rows, score completed rows, then clear both temporary flags. Do not
reset accumulated assessed or score values unless explicitly asked to
discard prior observations.