Aiming　at　the　lack　of　functional　versatility　in　current　engineering　attachments,　an　innovative　clamp-shear-grab　integrated　attachment　(CSGI　attachment),　which　combines　clamping,　shearing,　and　grabbing　functions,　is　proposed　based　on　the　author＇s　US　patent.　The　variable　frame　four-bar　mechanism　and　separate　frame　design　ideas　are　adopted　in　mechanism　design,　and　the　attachment＇s　motion　scheme　and　virtual　prototype　are　presented.　In　optimization,　the　optimization　objectives　of　the　CSGI　attachment　are　chosen　as　the　gripper　opening　size,　the　transmission　angle　of　the　clamp-shear　mechanism,　and　the　weight,　so　a　multi-objective　optimization　model　with　mechanism　dimension　and　topological　structure　is　established.　Next,　an　improved　nondominated　sorting　genetic　algorithm　(WINSGA-II)　is　proposed　by　improving　the　crossover　and　search　strategies,　while　its　effectiveness　is　tested　using　standard　indicators　and　data　sets.　Testing　results　show　that　the　WINSGA-II　algorithm　has　better　convergence,　population　diversity,　and　uniformity　than　the　nondominated　sorting　genetic　algorithm　(NSGA-II).　Subsequently,　the　WINSGA-II　algorithm　is　used　to　obtain　the　optimal　results　for　the　CSGI　attachment.　Finally,　the　engineering　prototype　is　made　and　tested　with　clamping,　shearing,　and　grabbing.　The　results　show　that　the　CSGI　attachment　can　achieve　the　intended　function.　Regarding　optimization,　the　gripper　opening　size　of　the　CSGI　attachment　is　expanded　by　24.2%,　and　the　weight　is　reduced　by　10.1%.　The　HV　and　Spacing　indicators　of　the　WINSGA-II　algorithm　are　2.7%　higher　and　29.6%　lower　than　those　of　the　NSGA-II　algorithm　for　this　problem.