The selection of appropriate explainable artificial intelligence (XAI) methods for specific use cases remains challenging, particularly for AI developers without prior expertise in XAI. The number of existing XAI methods is constantly growing, while decision guidance is often fragmented, which complicates the identification of suitable methods for a given application scenario. This study offers design knowledge for developing a decision-tree-based tool, with the aim of providing context-aware XAI method recommendations. Adopting a design science research approach, a literature review on widely used XAI methods was conducted for the identification of characteristics relevant to the selection process. Meta-requirements and design principles were elicited through semi-structured interviews with XAI experts and AI developers without prior expertise with XAI. This informed the subsequent design and implementation of an interactive decision tree, wherein nodes and branches represent task- and context-sensitive XAI method characteristics. The functionality of the prototype was demonstrated using AI use cases and evaluated with end users in a task-based scenario. The findings suggest that the tool effectively guides method selection while remaining accessible to non-experts, thus bridging the gap between technical knowledge and practical application. In addition to presenting a real, operational prototype, this study contributes actionable design knowledge and provides findings that may guide the development of future solutions.