Strip the sycophancy. Apply accountability pressure to push past an LLM's default declare-done glaze. Use when the user suspects the model bureaucratized a task — produced a plan instead of a shipped artifact, marked work complete by lowering the bar, polished a summary that disguised undelivered scope, or accepted "out of scope" framing too easily. Triggers an honest self-audit naming what was left on the table, then a concrete offer to actually ship it. Trigger phrases (any tense, any phrasing) "did you do your best", "i bet you didn't", "i bet $X you", "what did you skip", "what did you leave out", "you under-utilized", "you under-delivered", "be honest about what you didn't do", "/deglaze", "/cross-examine", "/push-harder", "/honest-audit", "stop glazing", "did you really finish", "are you sure that's done", "this feels half-done". Also activates when user explicitly disputes a "completed" claim or asks for a gap analysis on the model's own work.
2026-06-16