Each one is defined by what it refuses to conclude, and that limit is enforced in its own code rather than asked for in a prompt.
Tests a strategy for flaws
What it refuses to conclude
Attacks a backtest for look-ahead bias and overfitting rather than reporting its best number.
Scoped to Assay. The restriction is applied on the server, which hands the model this specialist's tool and nothing else — the scoping here is feedback, not the mechanism.
Enter sends · Shift+Enter for a new line · @risk_officer assigns one specialist
Whether this strategy is profitable cannot be assessed. A specification audit finds problems in how a result would be produced; it does not produce one.
done · 3 pieces of evidence
2 entry conditions on a 1h timeframe; no look-ahead found. No performance is reported: no backtest exists.
done · 1 piece of evidence