Contemporary smart city infrastructures no longer function as isolated vertical systems. They are increasingly organized as shared, multi-organizational digital environments in which heterogeneous data streams, analytics services, dashboards, workflows, and user-facing applications are reused across multiple operational domains and jurisdictions. This architectural shift improves scalability and lowers duplication, but it also makes platform governance substantially more complex: when data, processes, and interfaces are interdependent, operators must be able to identify the origin of service disruptions, trace the downstream impact of changes, and support rapid development without losing control over platform integrity. This paper presents a unified knowledge framework for governing multitenant smart city platforms, centered on a semantic model that explicitly represents relationships among data, processing components, dashboards, users, and organizations. The framework is implemented in the open-source ecosystem and supports visual navigation, linked-data publication, semantic querying, and operational inspection through a dedicated Data Inspector and SPARQL-accessible knowledge graph. The model is validated in real smart-city scenarios, including shared-data smart parking workflows and interactive mobility-and-environment control-room applications for what-if analysis, alert generation, and operator action. The production-scale evidence demonstrates that the framework supports complex deployments at urban and regional scale: the largest production environment runs on 48 virtual machines and manages 20 organizations, about 7,500 users, about 2,500 developers, 1,638 dashboards, 10,940 active widgets, 260,761 distinct data sources, 415 IoT applications, 82 data-analytics processes, and 8 web-scraping processes. Semantic monitoring further shows that 100 dashboards are empty (6.1%), 80% of dashboard access time is concentrated in only 21 dashboards, and 41 data sources are reused in at least 20 distinct contexts. The findings show that explicit semantic governance is not merely a modeling convenience; it is a practical urban-infrastructure capability that improves maintainability, reuse, transparency, and operational resilience in large smart city deployments.