The ENUBET ERC project (2016-2021) is studying a narrow band neutrino beam where lepton production can be monitored at single particle level in an instrumented decay tunnel. This would allow to measure $ u_{mu}$ and $ u_{e}$ cross sections with a precision improved by about one order of magnitude compared to present results. In this proceeding we describe a first realistic design of the hadron beamline based on a dipole coupled to a pair of quadrupole triplets along with the optimisation guidelines and the results of a simulation based on G4beamline. A static focusing design, though less efficient than a horn-based solution, results several times more efficient than originally expected. It works with slow proton extractions reducing drastically pile-up effects in the decay tunnel and it paves the way towards a time-tagged neutrino beam. On the other hand a horn-based transferline would ensure higher yields at the tunnel entrance. The first studies conducted at CERN to implement the synchronization between a few ms proton extraction and a horn pulse of 2-10 ms are also described.