Currently,　stretchable　strain-sensing　platforms　are　highly　desirable　for　the　full-range　and　accurate　detection　of　human　motions.　However,　it　is　also　a　grand　challenge　to　fabricate　strain　sensors　with　a　large　sensing　range,　high　sensitivity,　and　low　detection　limit.　Inspired　by　the　gradient　and　hierarchical　structures　in　bones,　a　high-performance　and　multifunctional　sensing　platform　with　a　gradient　component　distribution　was　reported　by　selecting　porous　carbon　polyhedrons　(PCP),　PCP　-welding-carbon　nanotubes　(PCP-w-CNTs),　and　CNTs　as　the　functional　nanomaterials　to　construct　a　novel　bone-mimicking　gradient　conductive　network　by　a　facile　layer-by　-layer　assembly,　and　elastic　poly(styrene-block-butadiene-block-styrene)　as　the　periosteum-like　coating　to　being　attached　to　conductive　networks.　Benefiting　from　the　gradient　structure　and　junction　welding　by　manipulating　the　vertical　component　distribution　of　conductive　fillers　and　in　situ　deposition,　the　strain　sensors　could　detect　small　and　large　deformations　with　an　ultralow　detection　limit　(-0.01%　strain),　a　large　sensing　region　(-660%　strain),　high　sensitivity　(-5797.83),　fast　response/recovery　time　(-180/180　ms),　and　excellent　robustness　(-1000　stretch-release　cycles).　Moreover,　the　composites　can　sense　liquids　due　to　their　different　solubility　parameters.