physical-ai-neural-reconstruction

To route NuRec tasks, describe your 3D reconstruction goal and the router directs you to the correct upstream sibling skill (physical-ai-datasets, ncore, nre, asset-harvester, nurec-fixer) for execution based on your workflow stage.

The best way to access canonical NuRec recipes is through the router, which guides you to fetch and reference the canonical upstream repository and its per-sibling SKILL.md without duplicating upstream logic.

You cannot perform data conversion or training directly within the router, as it has explicit safety boundaries preventing mutating actions; it only directs you to the authoritative upstream steps for these tasks.

Yes, the NuRec router supports a defined scope including dataset access, data conversion, training, and rendering workflows to coordinate your end-to-end 3D reconstruction process.

The router directs you to upstream sibling skills because it relies on the upstream skills index to prevent mutating actions within the router, ensuring you execute canonical recipes without duplicating upstream logic.

You need to use nurec-fixer or asset-harvester when the router identifies your user intent and NuRec workflow stage as requiring fusion or specific dataset handling, directing you to the appropriate upstream skill.