<p>This paper achieves prescribed performance control of multiagent systems (MASs) with multiple control directions using the backstepping technique. First, a distributed observer is constructed to transform the consensus problem into a tracking problem. Then, a tracking error-based P-like function is established to ensure that consensus errors evolve within the predefined region. Meanwhile, by integrating a scaling function, the constraint on initial conditions is relaxed. Unlike most existing results, the proposed strategy does not require the multiple control directions of agents to be identical. To address the challenge, a subtle Nussbaum function is introduced and a recursive control structure is developed. In addition, actuator faults with time-varying control directions are also studied, where unknown actuator effectiveness faults are treated as a composite unknown function and solved by the Nussbaum function, while additive faults are estimated and compensated by introducing a new variable function. The stability of the closed-loop system is analyzed based on the Lyapunov stability theorem with a contradictory discussion. Finally, simulation results demonstrate the effectiveness of the proposed methods.</p>