The review card is the last screen before a test runs, and it exists for one reason: to let you catch a misreading before it costs you a run rather than after.
Read the tags down the left
Each line is tagged — pair, when, buy, sell, stop, target, risk. The tags are a fixed column so the card reads as a specification rather than as prose, and so a missing line is visible by its absence. No "stop" line means no stop loss was parsed.
The bold part of each line is the part we extracted from your description. The rest is scaffolding. If the bold text is wrong, the strategy is wrong.
The blocks line
Underneath, the card says what the strategy was built from: how many blocks, and which kinds. This is generated from the strategy itself, not written by the model, so it cannot describe a block the strategy does not contain.
"Show the technical version" opens the raw structured form. You never need it, but it is there because a claim you cannot inspect is a claim you have to take on trust.
Changing something
Say what is wrong in the chat. The canvas rebuilds around the change and the card is rewritten. Corrections belong where every other instruction went, so there is no separate edit mode to learn.