IROS Robotics Reproducibility
Strengthens IROS conference papers by categorizing claims into rerunnable simulation code and auditable real-robot logs within strict page limits.
IROS Robotics Reproducibility is a specialized Claude Code skill tailored for robotics researchers preparing submissions for IEEE/RSJ IROS and similar robotics venues. Unlike standard machine learning papers, robotics research involves real-world hardware constraints that prevent readers from simply rerunning experiments. This skill guides authors to clearly partition claims into rerunnable assets (simulation scripts, seeded configurations, deterministic environments) and auditable evidence (hardware specification ledgers, calibration protocols, sensor firmware revisions, and trial logs). It ensures honest code/data availability statements, transparent sim-to-real gap reporting, and optimal organization directly inside the conference paper body without relying on supplementary PDFs.