Analyze Family Health History and Risks

Family Health Analyzer is an agent skill that analyzes family medical history and household health data. It covers five areas: genetic risk assessment, recognition of familial disease patterns, analysis of health problems shared across the household, personalized prevention advice, and visual report generation. It is written in Chinese and works with the tools Read, Write, Grep and Glob.

The analysis runs in eight steps. It first identifies the request type (family history analysis, genetic risk assessment, family health trends or a family health report), then reads the main data file data/family-health-tracker.json together with data/hypertension-tracker.json, data/diabetes-tracker.json and data/profile.json. Data is validated for relationship completeness, plausible ages and internal consistency. Pattern recognition then looks at familial clustering, inheritance patterns and early-onset cases, which it defines as typically under 50 years old.

Risk is computed as a weighted score: affected first-degree relatives count 0.4, early-onset cases 0.3 and familial clustering 0.3. Scores of 70% or more are high risk, 40% to 69% medium and below 40% low. Prevention advice falls into three groups: screening (periodic checks), lifestyle (diet, exercise, sleep) and medical care (when to see a doctor and which specialist to consult). A sample recommendation is structured data with a category, an action such as periodic blood pressure monitoring, a frequency of three times a week, a start age of 35 and a priority of high.

Use it when the task involves family medical history, genetic risk, family-level health patterns, family health reports, detection of diseases that cluster in families, or combining the health data of several family members into a trend or risk view. Typical requests are for a family health report, a family history analysis, a genetic risk assessment or family health trends, or running /family report or /family risk.

Do not use it to diagnose genetic disease, to predict whether a particular person will fall ill, or to recommend a specific treatment. Those are explicit safety boundaries of the skill. It is also not a substitute for environment-specific validation, testing or expert review, and it should stop and ask for clarification when required inputs, permissions, safety boundaries or success criteria are missing.

Inputs are the JSON tracker files listed above, plus the user's request or one of the two slash commands. Outputs come in two forms: a concise text report printed on the command line, and a full HTML report. The HTML report has four components: a family tree drawn as an ECharts tree chart, a genetic risk heatmap, a disease distribution pie chart and a timeline of prevention advice.

Every analysis must end with a disclaimer stating four points: the analysis rests on statistics from family history and is for reference only, the genetic risk assessment does not predict individual disease, all medical decisions should be taken with a qualified physician, and genetic counseling questions belong with a professional genetic counselor. Results should also state their uncertainty clearly.

It builds on existing modules: it reads hypertension management data and diabetes management data, and it links them with medication records. Maintainer credit in the skill goes to WellAlly Tech, version 1.0, last updated 2025-01-08.