Multistability in a dynamical system has attracted great attention recently for its complex and unexpected states. Since in most chaotic systems coexisting attractors reside in their own individual basin of attraction with a fractal structure, it becomes a challenge to choose correct initial conditions to obtain desired dynamics. Selecting typical dynamics as the basic components in a dynamical sequence and then arranging them in the phase space in a desired order make the multistability transparent and controllable in the domain of initial conditions; thereafter, one can identify an attractor according to its initial sequence. Dynamics editing provides an effective technique to select typical attractors under different system parameters to form a flexible sequence in the phase space, which shows great potential for chaos-based secure communications.