3C Quality Assistant

Problem

Quality work in 3C hardware programs often stalls on two practical issues: templates become empty frameworks when they lack key inputs, scoring rationale, and closure checks; and terminology routing gets ambiguous, for example between FMEA, PFMEA, and DFMEA, supplier onboarding versus supplier audit, or incoming material defects versus line defects. This skill treats quality tasks as executable sub-workflows: collect required inputs, preview the draft, then generate reviewable reports instead of producing unchecked long-form text.

How It Works

• Routing and intent disambiguation: uses trigger words, exclusion rules, and decision-tree pseudocode to separate scenarios such as S01 quality planning, S02 DFMEA, S08 8D, S11 complaints, S05 supplier onboarding, and S06 supplier audit.
• Quality calculations and judgments: supports RPN = S × O × D ranking, CPK process capability, AQL sampling decisions, COQ cost breakdown, SPC control-chart trends, and defect-rate Top 3 analysis.
• Document generation: covers FMEA worksheets, 8D reports, complaint records, supplier onboarding assessments, PVT release reviews, and quality weekly/monthly reports, with output variants such as Markdown, Excel, PDF, PPT, and quick-reference formats.
• Closed-loop follow-up: outputs from S02 DFMEA, S08 8D, and S11 complaints can be passed into FMEA updates, NUDD management, RMA, and CAPA flows, creating a traceable quality tracking chain.

Boundaries

It is best suited for small-sample, demo-level analysis and structured drafting, such as ranking given defect data by RPN or producing a first 8D draft. Large production datasets, real-time SPC collection, MES/PLM workflows, certification test conclusions, and field evidence still require professional systems and on-site validation. When critical inputs are missing, the skill asks for clarification rather than substituting engineering judgment.