In this article, we discuss the automatic inference of sufficient preconditions by abstract interpretation and sketch the construction of an under-approximating backward analysis. We focus on numeric properties of variables and revisit three classic numeric abstract domains: intervals, octagons, and polyhedra, with new under-approximating backward transfer functions, including the support for non-deterministic expressions, as well as lower widenings to handle loops. We show that effective under-approximation is possible natively in these domains without necessarily resorting to disjunctive completion nor domain complementation. Applications include the derivation of sufficient conditions for a program to never step outside an envelope of safe states, or dually to force it to eventually fail. We built a proof-of-concept prototype implementation and tried it on simple examples. Our construction and our implementation are very preliminary and mostly untried; our hope is to convince the reader that this constitutes a worthy avenue of research.