To account for cytoskeleton remodeling as well as smooth muscle length adaptation, here we represent the cytoskeleton as a 2-D network of links (contractile filaments or stress fibers) that connect nodes (dense plaques or focal adhesions). The network evolves in continuous turnover with probabilities of link formation and dissolution. The probability of link formation increases with the available fraction of contractile units, increases with the degree of network activation, and decreases with increasing distance between nodes, d, as 1/d^s, where s controls the distribution of link lengths. The probability of link dissolution decays with time to mimic progressive cytoskeleton stabilization. We computed network force (F) as the vector summation of link forces exerted at all nodes, unloaded shortening velocity (V) as being proportional to the average link length, and network compliance (C) as the change in network length per change in elastic force. Imposed deformation caused F to decrease transiently and then recover dynamically; recovery ability decreased with increasing time after activation, mimicking observed biological behavior. Isometric contractions showed small sensitivity of F to network length, thus maintaining high force over a wide range of lengths; V and C increased with increasing length. In these behaviors, link length regulation, as described by the parameter s, was found to be crucial. Concerning length adaptation, all phenomena reported thus far in the literature were captured by this extremely simple network model.