The task of decomposing a problem into sub-problems to build a solution, also formalized as planning in prior work, is a key skill for programming expertise. Improving the decomposition and planning skills of novices is shown to be a challenging goal for educators. While there are many planning methods employed by experts, studies show that experts tend to identify subtasks that can be solved via common code patterns. Many works have explored explicit instruction on common patterns in programs (i.e. programming plans) to support students in acquiring these skills. However, recent work implies that students may need additional support to fully benefit from such interventions. Moreover, tools like large language models (LLMs) are rapidly improving, and the skill of decomposing complex projects into smaller subtasks is increasingly relevant. This panel aims to bring computing education researchers together to discuss the main challenges around teaching decomposition and planning using common patterns, the crucial factors for designing instruction for teaching these concepts, and the impact evolving technology like LLMs can have on these developments.