This proposal introduces a novel methodology for constructing causal models that seamlessly integrate with the multi-layered ontological framework of critical realism, aimed at enhancing understanding and intervention in complex social phenomena. Developed as part of a PhD project on the causal mechanisms leading to the diagnosis of borderline personality disorder (BPD), this method employs directed acyclic graphs (DAG) to visually map and analyze causal pathways, facilitating both scholarly discourse and practical application in mental health research. The methodology is designed to be inherently interdisciplinary and adaptable across various domains, making it applicable to a broad range of disciplines beyond mental health. By leveraging the principles of grounded theory and expert knowledge elicitation along with existing literature, the approach ensures a rigorous empirical foundation, while the iterative nature of the model-building process supports continuous refinement and validation of the causal model. Central to this methodology is its commitment to collaboration and openness. Inspired by the model of free and open source software, it utilizes Git-based principles to allow for an open-source framework where models can be shared, modified, and enhanced by a community of researchers. This not only democratizes the research process but also promotes transparency in scientific inquiry. The application of this methodology to the study of BPD serves as a pivotal case study, demonstrating its potential to uncover nuanced causal relationships and foster a deeper understanding of psychiatric diagnoses within the critical realist paradigm.