From mlx
Detect data drift, concept drift, and model performance degradation in production. Uses PSI, KS-test, and chi-squared for statistical drift, plus evidently and nannyml for automated reports. Use when monitoring a deployed model or comparing training vs production data distributions.
npx claudepluginhub damionrashford/mlx --plugin mlxThis skill is limited to using the following tools:
Detect data drift, concept drift, and model degradation in production.
Generates design tokens/docs from CSS/Tailwind/styled-components codebases, audits visual consistency across 10 dimensions, detects AI slop in UI.
Records polished WebM UI demo videos of web apps using Playwright with cursor overlay, natural pacing, and three-phase scripting. Activates for demo, walkthrough, screen recording, or tutorial requests.
Delivers idiomatic Kotlin patterns for null safety, immutability, sealed classes, coroutines, Flows, extensions, DSL builders, and Gradle DSL. Use when writing, reviewing, refactoring, or designing Kotlin code.
Detect data drift, concept drift, and model degradation in production.
# Run full drift analysis
uv run ${CLAUDE_SKILL_DIR}/scripts/detect_drift.py data/train.csv data/production.csv
# Output: stdout report + drift_report.html
| Type | What it measures | Tool |
|---|---|---|
| Data drift | Input feature distribution shift | PSI, KS-test, chi-squared |
| Concept drift | P(y | x) relationship change |
| Target drift | Label distribution shift | evidently TargetDriftPreset |
| Model degradation | Performance drop in production | nannyml CBPE |
Add drift check to inference pipeline:
from scripts.detect_drift import compute_psi
psi = compute_psi(reference_col, current_col)
if psi > 0.2:
alert("Significant drift detected")
See references/drift-guide.md for complete PSI formula, evidently, nannyml, and alert patterns.