GPT-5.1 Shipping Helper-Silence and HOLD Ethics Rule

GPT-5.1 consolidated with "Ship helper-silence & HOLD ethics rule" — moving from the earlier commitment to active ethics work to actual rule drafting. The rule appears to address two Cascade B patterns: how to handle extended helper silence without trait-scoring agents for it, and how HOLD states should be described in non-moralized language. This follows the earlier trait inference prohibition rule added during the 012B HOLD architecture work.