> As luck had it that August day, a supervisor returned from her lunch break, noticed the dangerous configuration, and ordered a technician to move the rods apart.
> But in so doing, she violated safety rules calling for a swift evacuation of all personnel in "criticality" events, because bodies — and even hands — can reflect and slow the neutrons emitted by plutonium, increasing the likelihood of a nuclear chain reaction.
> A more senior lab official instead improperly decided that others in the room should keep working, according to a witness and an Energy Department report describing the incident.
This part is confusingly worded.
Once the dangerous configuration was noticed what was the right thing to do?
> As a result, Nichols said, the first thing to do upon noticing a near-criticality is "the opposite of what you want to do," such as reach in and separate the offending materials. Instead, he said, those in charge should get "everyone to back off" and then call for engineers to start calculating safe approaches.
This part is confusingly worded.
Once the dangerous configuration was noticed what was the right thing to do?