/rigorous-evaluation
Use when evaluating or reporting model performance and choosing metrics, thresholds, and plots that FIT the problem type. Picks the right metrics per task (binary uses ROC-AUC and PR-AUC; multiclass uses macro F1 and a confusion matrix; detection uses mAP; segmentation uses Dice
$ npx -y skills add mxslr/mlcraft --skill rigorous-evaluation --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.
- You can call itInvoke it directly when you want it.
- Slash command
/rigorous-evaluation
Context preview
The summary Claude sees to decide when to auto-load this skill.
Use when evaluating or reporting model performance and choosing metrics, thresholds, and plots that FIT the problem type. Picks the right metrics per task (binary uses ROC-AUC and PR-AUC; multiclass uses macro F1 and a confusion matrix; detection uses mAP; segmentation uses Dice
A research-first AI/ML research-engineer workflow for Claude Code
Other skills on mlcraft.
- /accuracy-improvement-loop
Use when an existing model's results are disappointing and the user wants higher accuracy - 'accuracy is still too low', 'improve/boost the model', 'why is it bad', 'make it better than the paper', 'combine methods to beat SOTA'. Diagnoses the real bottleneck first (error
Open skill - /data-rigor-and-leakage
Use BEFORE training any model, to build correct train/val/test splits and hunt data leakage - the #1 cause of fake-high accuracy. Covers group/patient/subject splits, temporal splits, official-benchmark splits, label correctness, class balance, and preprocessing parity. Triggers
Open skill - /dataset-profiling
Use as the FIRST step of any ML task, before choosing a model, to inspect and understand the actual dataset. Works for a LOCAL dataset (Claude reads the files directly) and for a KAGGLE dataset (Claude cannot read /kaggle/input from your machine, so it emits a small profiling
Open skill - /deployment-explainability
Use when turning a trained model into something usable (an app, an inference function, a demo) or adding explainability. Selects the RIGHT interpretability method per problem: Grad-CAM for CNN image classification, predicted boxes for detection, mask overlays for segmentation,
Open skill - /domain-3d
Use for 3D and point cloud tasks: point cloud classification, 3D semantic or instance segmentation, 3D object detection from LiDAR, and depth or mesh analysis, including autonomous-driving and indoor-scan settings. Picks the right 3D backbone, representation, and scene-level
Open skill - /domain-anomaly-detection
Use for anomaly, outlier, and novelty detection across data types: industrial or visual defect detection, time-series or sensor anomalies and predictive maintenance, tabular fraud or intrusion detection, and out-of-distribution detection. Usually one-class or unsupervised. Picks
Open skill