12 soil-science skills forged from the DSM knowledge wiki: soil quality & health indexing (SQI/SHI — minimum data set, indicator scoring, weighting, aggregation, validation, SEM, fuzzy rules), fertility mapping, prediction uncertainty (AOA/LPD/Monte-Carlo), sampling design, ML for soil, ensemble ML (whether to combine models at all — the ambiguity ceiling, the weighting spectrum, cross-fitted combiners), end-to-end DSM workflow, map validation, RothC temporal modelling, hybrid process+ML SOC, SOC-stock MRV, and spatial statistics on areal data (Moran/LISA/BiLISA, SAR/SEM/SDM/SDEM, spillover impacts) with verified R and Python pipelines.