No user-submitted related servers found.
Scored across 8 tools
Each tool targets a distinct stage of the ML pipeline (train, sample input, per-model predict, metrics, selection, SHAP, export). The only mild overlap is predict_logistic_regression vs predict_knn, but the names and descriptions clearly distinguish which model each uses.
Most tools follow a verb_noun pattern (train_models, predict_knn, select_best_model, export_parquet_audit). A couple deviate into noun-heavy forms (confusion_matrix_metrics, shap_explain_selected), but the set remains readable and predictable.
8 tools is well-scoped for a focused ML agent, mapping cleanly onto the train/evaluate/select/explain/export lifecycle. No filler tools and no obvious missing step.
The pipeline covers training, per-model prediction, evaluation, model selection, explainability, and audit export. Minor gaps exist (e.g. no predict endpoint for arbitrary new input beyond the provided sample, no listing of saved artifacts), but these are workable around.