About this skill

Problem It Solves

Quality teams often struggle to turn defect logs, root-cause notes, and process measurements into consistent visual analyses. Manual charting makes it hard to compare Pareto rankings, control limits, or distribution shapes across teams. This skill organizes common QC methods into a repeatable workflow: define the analysis goal, prepare the data, generate the chart, and export results for review. It is useful when converting records like cosmetic defects, dimension deviation, or assembly failures into a Pareto chart, breaking down “high defect rate” into an Ishikawa diagram, or checking whether batch means stay within control limits.

How It Works

The skill covers five chart types:
- Pareto: Sorts defect categories by count and highlights the key issues that make up roughly 80% of the total.
- Ishikawa: Expands possible causes across dimensions such as man, machine, material, method, environment, and measurement.
- Histogram: Visualizes the distribution of one numeric dataset to support process-capability review.
- Control chart: Accepts subgroup data and subgroup size, then plots UCL, CL, LCL, and data points.
- Scatter: Plots two numeric variables to inspect correlation strength and direction.

A typical run starts with the analysis goal, followed by data input or a preset template. The script scripts/qc_chart_generator.py generates the chart, templates live in assets/templates/, and references/qc_guide.md covers operation steps and interpretation. Charts default to PNG, with PDF or SVG available through parameters. Data and analysis results can be exported as CSV or JSON for archiving or downstream analysis.

Scope And Cautions

Input data must match the JSON format expected by the script, especially subgroup structure for control charts. UCL and LCL can be calculated automatically or specified manually; if the business rule already defines limits, pass them explicitly. For large datasets, batch processing is recommended to avoid performance issues. This skill is best treated as a quality-analysis and visualization aid rather than a production data extractor, statistical modeling tool, or formal audit conclusion.

Screenshots

Use Cases

A quality engineer has one week of defect categories and counts and needs a Pareto chart to identify the top issues making up about 80% of defects.
A process engineer is investigating high product defect rates and needs an Ishikawa diagram across man, machine, material, method, environment, and measurement.
A production engineer enters subgroup means and counts for 25 shifts and needs an X-bar control chart showing UCL, CL, LCL, and out-of-limit points.
A data analyst prepares two numeric process variables and needs a scatter plot to judge correlation strength and direction, then export CSV.

Best For

Quality engineers who run weekly defect reviews and need to turn category counts into a Pareto chart.
Process or quality engineers who need to locate defect root causes using an Ishikawa diagram across man, machine, material, method, environment, and measurement.
Production engineers who monitor process stability and need control charts from subgroup data to detect out-of-limit points.
Data analysts who archive quality results and need chart data and findings exported as CSV or JSON.