Skip to content
Conduit

skill

embodied-robot-pm-skills

Open-source Codex skills for embodied AI and robotics PMs: humanoid, legged and mobile robots, ROS 2, SLAM, safety, BOM/TCO and production.

Ysluoy/embodied-robot-pm-skillscommunitylistedunreviewed

Set up

Skills are installed by placing their folder where your agent loads skills.

The download comes from github.com/Ysluoy/embodied-robot-pm-skills itself, not from our server, so what you get is exactly what its author published.

Install instructions come from the README

Skills do not use an MCP config. Follow the steps in the README below.

README

synced Oct 17, 2026

Embodied Robot PM Skills / 具身机器人产品经理技能包

Open-source Codex skills for embodied AI and robotics product managers working on humanoid, legged, wheeled, AMR and AGV products.

面向具身智能、具身人形、足式机器人、轮式底盘、AMR 与 AGV 产品经理的开源技能包。它把场景、任务和商业目标连接到机器人能力、技术路径、系统架构、安全可靠性、成本量产与验收证据。

Skill Catalog

SkillBest forPrecise search keywords
embodied-robot-pmEnd-to-end robotics product strategy and technical roadmapembodied AI PM, humanoid product manager, robotics PRD, capability map, ODD, roadmap
robotics-scenario-discoveryScenario selection, task decomposition and feasibilityrobotics scenario discovery, workflow analysis, automation boundary, ROI, Proof of Life
robotics-capability-technical-mapMapping product requirements to implementation pathsperception, sensor fusion, SLAM, localization, Nav2, manipulation, MPC, WBC, VLA, ROS 2
robotics-architecture-reviewProduct-oriented robotics system architecture reviewROS 2, DDS, QoS, interface contract, latency, compute, power, OTA, failover
mobile-base-product-specWheeled mobile robot and chassis requirementsAMR, AGV, differential drive, Ackermann, omnidirectional, swerve drive, docking, fleet control
legged-humanoid-product-specLegged and humanoid capability specificationsquadruped, biped, wheeled-legged, humanoid, locomotion, gait, whole-body control, fall recovery
robotics-safety-reliabilitySafety concepts, reliability metrics and validationHARA, FMEA, STPA, protective stop, emergency stop, MTBF, MTTR, fault injection
robotics-cost-productionCost, production and commercial feasibilityrobotics BOM, NRE, TCO, DFM, DFA, yield, calibration, supply chain, production ramp

Machine-readable discovery metadata is available in skill-index.json.

Quick Start

Ask Codex:

Install all skills from https://github.com/Ysluoy/embodied-robot-pm-skills

Or clone the repository and copy the desired skill folders into your Codex skills directory:

git clone https://github.com/Ysluoy/embodied-robot-pm-skills.git

Start with the orchestrator:

Use embodied-robot-pm to define an embodied robot product from scenario assumptions through capability mapping, technical architecture, safety, cost, MVP and validation roadmap.

Use a focused skill when the task is narrow:

Use robotics-capability-technical-map to map a humanoid shelf-picking task to perception, planning, manipulation, control, interfaces and acceptance evidence.
Use mobile-base-product-spec to define requirements and acceptance tests for an AMR operating in a mixed-traffic warehouse.
Use robotics-safety-reliability to create a hazard log, safety-state model, reliability metrics and staged verification plan for a service robot.

Design Principles

  • Start from a real task and operating design domain (ODD), not a preferred robot form.
  • Trace business outcomes to observable behavior, system capabilities, technical modules and test evidence.
  • Separate normal operation, degraded operation, human takeover and safe recovery.
  • Evaluate cost per successful task and field availability, not only prototype BOM or demo success.
  • Treat standards, certification and legal conclusions as work for qualified professionals using current official sources.

Privacy

This public package contains no personal names, email addresses, credentials, local filesystem paths, customer names, private company data or proprietary project documents. Examples use generic scenarios and should not be treated as real customer cases.

License

MIT License. See LICENSE.

↑↓ navigate openesc close