Take the Person of the first item, the Action of the second, and the Object of the third, and stage them as one scene.

* Each card owns a full Person/Action/Object — the composite takes the Person of card 1, Action of card 2, Object of card 3, and drops the rest. *

Given three items in order — call them 1, 2, 3 — you use exactly one slot from each:

• Person from item 1
• Action from item 2
• Object from item 3

So three cards A♠ · 7♥ · K♦ become Bond (A♠'s person) surfing (7♥'s action) on a cake (K♦'s object). One image, one place, three items — in order.

Go deeper:

• Mnemonic peg system (Wikipedia) — grounds the 00–99 P/A/O structure the rule draws on.
• Nelson Dellis teaches Bill Weir a deck — a champion demonstrating person-of-1, action-of-2, object-of-3.

Read the slots in fixed order: the person names item 1, the action names item 2, the object names item 3.

Decoding is the exact inverse of encoding. You see "Bond surfing on a cake" and ask, in order:

1. Whose person is Bond? → item 1 (A♠).
2. Whose action is surfing? → item 2 (7♥).
3. Whose object is a cake? → item 3 (K♦).

Because every person/action/object belongs to exactly one item, each slot resolves to one answer — and the fixed order gives you the sequence.

Go deeper:

• Nelson Dellis teaches Bill Weir a deck — shows reading a meshed scene back into card order.