idea-discovery-robot

Automate robotics idea discovery with literature surveys, idea generation, and novelty checks.

Updated Apr 18, 2026

Search Tags:#benchmark #robotics #research-review #embodied-ai #novelty-check #idea-discovery #sim-first

Ready to run

npx skills add https://github.com/THUFanZd/Sewed_pipeline --skill idea-discovery-robot-thufanzd

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: idea-discovery-robot
Source: https://github.com/THUFanZd/Sewed_pipeline/tree/main/.agents/skills/idea-discovery-robot
Command: npx skills add https://github.com/THUFanZd/Sewed_pipeline --skill idea-discovery-robot-thufanzd

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The Robotics Idea Discovery workflow frames broad robotics directions into concrete, benchmark-grounded, simulation-first ideas, reducing manual framing and literature synthesis time for embodied AI projects.

Core Features & Use Cases

Orchestrates four sub-skills: robotics literature survey, idea generation, novelty check, and critical review, all grounded in robotics constraints (embodiment, task family, observation/action interfaces, sim2real considerations).
Produces a structured Robotics Problem Frame to guide evaluation, pilot design, and benchmarking.
Suitable for early-stage project scoping, benchmark selection, and evaluation planning for manipulation, locomotion, navigation, drones, or humanoid systems.

Quick Start

Start with a robotics direction and run the full pipeline to generate top sim-first ideas and a novelty-check plan.

Frequently Asked Questions about idea-discovery-robot

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I generate research ideas for embodied AI tasks like manipulation and locomotion?▼

To generate embodied AI ideas, you can use a robotics idea discovery workflow that frames problems, surveys literature, generates ideas, and verifies novelty against simulation benchmarks. This automates framing and literature synthesis for manipulation and locomotion tasks.

What is the best way to check if a robotics research idea is novel?▼

Checking robotics idea novelty involves running an automated novelty-check phase that evaluates generated concepts against existing literature. This verifies that your proposed manipulation or navigation approach has not been previously explored.

Can I use a sim-first approach to evaluate drone and humanoid navigation benchmarks?▼

Yes, you can use a sim-first evaluation approach for drone and humanoid navigation benchmarks. The workflow grounds generated ideas in simulation first, ensuring evaluation planning aligns with robotics constraints and specific task families.

How do I start framing a broad robotics direction into a concrete benchmark problem?▼

Framing a broad robotics direction starts by defining a structured Robotics Problem Frame that specifies embodiment, task family, and observation/action interfaces. This reduces manual scoping time and guides pilot design and benchmark selection.

Does the robotics idea generation workflow progress automatically if I do not provide input?▼

Yes, the robotics idea generation workflow features automatic progression if the user is unresponsive. It seamlessly moves through literature survey, idea creation, novelty check, and research review phases without requiring continuous manual input.