dmaic-measure

Establish quantitative process baselines with validated measurement systems in DMAIC projects.

1|Updated Mar 25, 2026
One-click install
npx skills add https://github.com/NestorMonroy/thyrox --skill dmaic-measure-nestormonroy
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: dmaic-measure
Source: https://github.com/NestorMonroy/thyrox/tree/main/.claude/skills/dmaic-measure
Command: npx skills add https://github.com/NestorMonroy/thyrox --skill dmaic-measure-nestormonroy

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve? Teams often jump into root-cause analysis without trustworthy data, producing conclusions built on unreliable measurements. This Skill guides the Measure phase of a DMAIC project so you define a measurement plan, validate the measurement system (MSA), and compute a defensible baseline (DPU, DPMO, Sigma Level, Cp/Cpk) before moving to Analyze. ## Core Features & Use Cases - Measurement planning and process mapping: Build a measurement plan from CTQs, supported by process maps, Value Stream Maps, and swim lane diagrams to identify where and what to measure. - Measurement System Analysis: Apply Gauge R&R acceptance thresholds for continuous data and Cohen's Kappa for attribute data, with decision tables for accept, conditional, or reject outcomes. - Baseline metrics and capability analysis: Calculate DPU, DPMO, Sigma Level (with the 1.5σ shift convention), and Cp/Cpk/Pp/Ppk, including a Python script that computes capability indices from CSV data with specification limits. - Use Case: A quality engineer needs to baseline defect rates on a production line. The Skill walks them through stratified sampling, validates inspector agreement with Kappa, computes DPMO and Sigma Level, and produces a structured Measure artifact ready for the Analyze tollgate. ## Quick Start Ask the assistant to run the DMAIC Measure phase for your approved project charter, defining the measurement plan, validating the measurement system, and calculating the Sigma Level baseline.

Frequently Asked Questions about dmaic-measure

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I calculate Sigma Level from DPMO in Six Sigma?▼

Compute DPU as total defects divided by total units, then DPMO as DPU divided by opportunities per unit times one million. Convert DPMO to Sigma Level using the standard Six Sigma table, which already incorporates the 1.5 sigma long-term shift convention.

How do I run a Gauge R&R study for measurement system analysis?▼

Use 2-3 operators measuring 10 representative parts with 2-3 randomized replicates each. Compute repeatability from within-cell ranges and reproducibility from operator averages, then evaluate %GR&R: under 10% is acceptable, 10-30% is conditional, and 30% or more requires fixing the measurement system.

What is the difference between Cp and Cpk in process capability?▼

Cp measures potential capability as (USL - LSL) / 6σ assuming a centered process, while Cpk accounts for actual centering using the minimum distance from the mean to either spec limit. A high Cp with low Cpk indicates a low-variability but off-center process.

When should I use Cohen's Kappa instead of Gauge R&R?▼

Use Cohen's Kappa when your CTQ data is categorical or attribute-based, such as pass/fail or defect category classifications. Kappa above 0.9 indicates excellent agreement, 0.7-0.9 is acceptable, and below 0.7 means classification criteria must be redefined before trusting the data.

Can I calculate Cpk if my process is not in statistical control?▼

No, Cp and Cpk are only valid when the process is in statistical control, verified with control charts. If special causes are present, eliminate them first, because capability indices computed on an unstable process are meaningless.

What are the limitations of the DMAIC Measure phase approach?▼

The Skill guides the measurement process but does not include statistical software like Minitab or R for advanced analysis. Processes with severe non-normality or complex correlations may require a Black Belt or statistician, and historical data without retrospective MSA carries documented uncertainty.