Development of sufficiently optimized software coding Contracts with insignificant association from developers directs at lowering the setbacks in abiding by the input specifications of the software, delivering systematic testing of several real life software. The associated framework put forward primarily speculates dependency aspects signifying both behavior and semantic attributes well-defined as conditions in a Decision tree as Automated Contracts for conformance with input specs. The contracts are perfected based on their feasibility with a modified form of Ant Colony Optimization algorithm motivated by pursuing behavior of ant colonies of real world, with considerable comprehensiveness in the Fault identification potential. This research focuses on the decision tree classifiers are J48, RandomForest classifiers have same precision value which is 0.96 of precision value. RandomTree and REPTree classifiers have same precision value which is 0.95 of precision value, HoeffdingTree classifier is 0.82 of precision, DecisionStump is zero precision value. The J48, RandomForest classifiers have same recall value which is 0.96 of recall value. RandomTree and REPTree classifiers have same recall value which is producing 0.95 of recall value, HoeffdingTree classifier is having 0.76 of recall value, DecisionStump is having zero precision value.The RandomForest is having highest ROC value whichi is 1. The J48 classifier , REPTree classifier have 0.99 ROC value. HoeffdingTree classifier is 0.98 of recall value, RandomTree classifier has 0.97 ROC value and finally the lowest ROC value is produced by DecisionStump Decision classifier.The RandomForest, RandomTree, REPTree classifiers are having highest PRC Area value which is 0.99. The J48 classifier is having 0.92 PRC Area, HoeffdingTree classifier is 0.85 of PRC Area, and DecisioStump classifier is having 0.23 PRC Area value. Among the set of decision classifiers, the J48 decision classifier is producing the best results based on the accuracy, precision, recall, ROC and PRC area values which are compare with other models.