The paper methodologically shows the identity of the mathematical problem of graph path searching with the technical problem of searching for various defects in software, in particular, bugs and undocumented features. The graph model of software functioning, which became the basis for the presented methodology, is briefly described. New research directions based on graph theory problems, which have not been previously used to search for defects in software, are stated.

Keywords: graph model, software, graph pathfinding, breadth-first search, meet-in-the-middle method, malicious software