Abstract:Modern software deployment is no longer a straightforward task. It now weaves through layers of networking, storage, and compute resources, all bound by intricate interdependencies. Ensuring the correctness and reliability of systems after deployment requires more than just sequencing steps and instead calls for precise modeling of dependencies and the generation of executable plans that advance in order from initial to terminal components. Yet current low-code platforms and mainstream deployment tools offer limited support for formal modeling and automation, posing a key obstacle to flexibility and reliability. This study addresses these challenges by adopting a formal model known as the unified structure, which models component dependencies and evaluates deployment feasibility in cloud environments. Building on this, this study proposes an automated approach to generate executable installation plans and develops a prototype tool to enable automated deployment. Experiments validate the method’s effectiveness and correctness.