The concept appears elegant and simple. The idea is ingenious to direct players towards algorithmic thinking, yet Automachef seems to be cruelly lacking in consistency and precision. The game offers different levels, in which specific dishes must be served to customers: the goal is to automate the production chain, while taking care of the number of ingredients used and the energy consumed. The title offers several obstacles to spice up the levels; sometimes it's about building the line in a small space – the best levels in my opinion –, sometimes it's about not exceeding a certain energy consumption, other times it's about delivering on two different outputs, and so on. The main problem is that the progression curve is surprisingly shaky. Automachef is quite difficult to get to grips with, at first, as there are so many options to adjust to handle the various mechanical arms and other machines. Then, the title becomes a long sequence of execution: the levels do not pose any more problems and you just have to copy and paste pre-made circuits, before making some terminal adjustments. Indeed, as the constraints change only slightly, it is easy to imagine the ideal solution, for which it will unfortunately take an hour of construction. The bonus levels offer more interesting challenges, forcing the use of more complex machines; nevertheless, the difficulty is poorly calibrated and will put off players who have not mastered the MIPS assembly: a sad observation. To this, we must add the sometimes capricious controls. It is not always possible to target the different machines and the absence of intermediate zoom is very frustrating. If Automachef knows how to be gratifying, one cannot help but consider the sessions on this game as purges, when the mass of work to be done appears before us.