Video-mediated interpreting (VMI) can enhance access to language support in asylum reception, but its success relies on context-sensitive guidelines and training. This paper introduces a VMI training framework developed through the European Project EU-WEBPSI, which consolidates research and professional standards into minimum requirements for asylum and refugee contexts. Addressing technical, interactional, and ethical dimensions, it clarifies shared responsibilities among interpreters, service providers, and users. Building on this foundation, the paper presents training materials that promote collaboration and sustainable delivery under technological and resource constraints. Using semi-scripted simulations that recreate real challenges - such as embodied cues, repair, and triadic coordination - it illustrates how scaffolded learning can benefit interpreter trainers and practitioners. Overall, the framework offers a coherent, research-informed approach to preparing professionals for VMI, with wider applicability across public service interpreting settings and implications for codes of practice, trainer development, and ongoing evaluation.