Official medical records maintained by healthcare institutions provide essential clinical information, yet their scope often remains fragmented and limited to institutional contexts. Complementing these records with personal patient summaries and enriched electronic health records provides a more holistic view of individual health, integrating lifestyle, patient history, and self-reported data. This paper analyzes a prototype that combines graph databases with multi-agent infrastructure to enable knowledge graph construction and querying. The study explores the potential of agentic systems, which draw on the linguistic and reasoning capabilities of Large Language Models, the connective infrastructure of the Model Context Protocol, and the patient-centered orientation of Personal Health Knowledge Systems. Together, these components create intelligent frameworks that extend beyond conventional automation by combining probabilistic reasoning, advanced autonomy, decentralization, scalability, and adaptability in digital healthcare.