OpenRaiser/NanoResearch

This skill library orchestrates the full AI research lifecycle in a continuous two-loop architecture. An outer loop steers direction by synthesizing literature, generating hypotheses, and reflecting on results. An inner loop runs rapid experiments against clear optimization targets using generated Python code skeletons, distributed training, fine-tuning, and multi-cloud infrastructure. The library serves AI research scientists, ML research engineers, and autonomous research teams who need to accelerate ideation, execution, and publication without pausing for manual handoffs.