Summary: This paper uses detector-based mode information to address the \(\mathcal{H}_2\) state-feedback control problem for continuous-time Markov Jump Linear Systems (MJLS). The focus is on designing robust controllers capable of handling uncertainties affecting both the Markov process and its detector, a less explored scenario in the literature. The main contribution lies in presenting a new LMI-based algorithm within a suitable synthesis framework to deal with uncertain parameters. A key advantage is the absence of constraints on the optimization variables, even in cases involving structured controllers like decentralized control. To illustrate the effectiveness of the proposed technique, a numerical example based on a linearized model of an unmanned aircraft is provided, showcasing its superiority over existing approaches.