硬件设计 Super Skill。一个父技能内含 25 个子技能（mechanical / viewer / urdf / srdf / sdf / simulation / gcode / sendcutsend / parts-catalog / bambu-labs / pcb / electronics-bom / requirements-verification / actuator-sizing / pcb-mechanical-reliability / circuit-simulation / gait-optimization / motion-control / firmware / sim2real-calibration / integration / fea / wear-fatigue / mujoco-simulation / robot-dog-digital-twin）， 覆盖机械建模 → 网页预览 → 机器人描述 → 动力学仿真 → 制造出工 → 电子域全链路。 触发词：build123d、CAD建模、做一个零件、参数化设计、装配、URDF、机器人描述、 MoveIt、Gazebo、SDF、动力学仿真、pybullet、跌落测试、步态仿真、headless sim、站得稳吗、 切片、3D打印、激光切割、SendCutSend、Bambu、找现成零件、 STEP、网页预览、分享链接、PCB、tscircuit、代码写PCB、Gerber、嘉立创、JLCPCB、下单打板、电子BOM、 需求合同、验证矩阵、执行器选型、扭矩裕量、热裕量、PCB刚度、PCB挠曲、连接器受力、 电源预算、电路保护、热风险、步态优化、行走算法、IK、FK、运动控制、轨迹生成、固件、嵌入式、CAN、校准、sim2real、仿真实机对齐、整机集成、bring-up、HIL、FEA、结构强度、刚度、磨损、疲劳、轴承寿命、维护周期、MuJoCo、MJCF、高保真动力学、数字孪生、虚拟样机、机械狗验证、 设计评分、实体样机 gate。 父级只做路由，详细能力在 skills/<name>/SKILL.md。
Robot-dog integration, bring-up, HIL planning, prototype readiness, hardware gate, assembly test, first power-on checklist, data capture, and physical test safety validation. Use this skill when the user asks whether a prototype is ready for assembly, power-on, HIL, bring-up, hardware testing, or real-world data collection after digital-twin gates.
Sim-to-real calibration skill for robot-dog virtual prototypes. Use this skill when comparing simulation logs against physical prototype logs, estimating friction/mass/latency/torque scale/contact parameter corrections, reducing sim-real gap, or generating parameter updates for MuJoCo/PyBullet/digital twin.
Robot-dog firmware dry-run planning and validation skill for MCU target, motor-control loop, CAN protocol, safety state machine, calibration contract, build manifest, and hardware bring-up gates. Use this skill when the user asks for embedded firmware, motor firmware, FOC loop, CAN frames, calibration, firmware project scaffolding, or safe pre-flash validation.
Electronics BOM and robot-dog component selection MVP. Use this skill when the user asks for electronic BOM, MCU/driver/encoder/power/connector selection, JLCPCB/LCSC-ready part candidates, curated robot electronics library, BOM rationale, availability checks, or component choices to feed pcb, circuit simulation, firmware, and digital-twin gates.
Executable robot-dog motion-control MVP for leg IK, gait phase generation, trajectory contracts, controller parameters, and simulation/firmware handoff. Use this skill when the user asks for FK/IK, inverse kinematics, gait generator, trot/walk/bound trajectories, controller parameters, motion-control code, or a trajectory to feed simulation, MuJoCo, or firmware.
MuJoCo-oriented robot-dog dynamics validation, MJCF scenario contracts, contact, terrain, actuator, stability, torque, slip, and disturbance checks for virtual prototypes. Use this skill when the user asks for MuJoCo, MJCF, high-fidelity legged-robot simulation, contact/friction validation, slope/step/drop/push scenarios, or serious gait simulation beyond PyBullet smoke tests.
Wear, fatigue, bearing life, harness bend, connector retention, foot pad wear, and maintenance interval checks for robot-dog and hardware virtual prototypes. Use this skill when the user asks whether gears, bearings, joints, feet, wires, connectors, or moving mechanical interfaces will wear out, fatigue, loosen, or block physical prototype release.