wellally-tech – skill Dữ liệu & Khoa học cho agent AI

WellAlly Digital Health Integration

Integrate multiple digital health data sources, connect to [WellAlly.tech](https://www.wellally.tech/) knowledge base, providing data import and knowledge reference for personal health management systems.

Detailed Guide

Read [the detailed guide](references/detailed-guide.md) before executing this skill. It retains the complete procedure and reference material. Treat its safety, prerequisites, and validation requirements as mandatory. For focused work, load the relevant sections; for end-to-end work, read the guide completely.

When to Use

• You need to import or normalize health data from sources like Apple Health, Fitbit, Oura, or CSV/JSON exports.
• You want to connect personal health data workflows to the WellAlly.tech knowledge base.
• The task involves data import, health-data management, or article recommendations driven by user health context.

Usage Instructions

Trigger Conditions

Use this skill when users mention the following scenarios:

Data Import:

• ✅ "Import my health data from Apple Health"
• ✅ "Connect my Fitbit device"
• ✅ "Sync my Oura Ring data"
• ✅ "Import CSV health data file"
• ✅ "How to import fitness tracker/smartwatch data"

Knowledge Base Query:

• ✅ "Articles about hypertension on WellAlly platform"
• ✅ "Recommend some health management reading materials"
• ✅ "Recommend articles based on my health data"
• ✅ "WellAlly knowledge base articles about sleep"
• ✅ "How to improve my blood pressure (check knowledge base)"

Data Management:

• ✅ "What health data sources do I have"
• ✅ "Integrate health data from different platforms"
• ✅ "View imported external data"

Execution Steps

Step 1: Identify User Intent

Determine what the user wants:

1. Import Data: Import data from external health platforms

2. Query Knowledge Base: Find [WellAlly.tech](https://www.wellally.tech/) related articles

3. Get Recommendations: Recommend articles based on health data

4. Data Management: View or manage imported external data

Step 2: Data Import Workflow

If user wants to import data:

2.1 Determine Data Source

const dataSource = identifySource(userInput);
// Possible returns: "apple-health", "fitbit", "oura", "generic-csv", "generic-json"

2.2 Read External Data

Use appropriate import script based on data source type:

// Apple Health
const appleHealthData = readAppleHealthExport(exportPath);

// Fitbit
const fitbitData = fetchFitbitData(dateRange);

// Oura Ring
const ouraData = fetchOuraData(dateRange);

// Generic CSV/JSON
const genericData = readGenericFile(filePath, mappingConfig);

2.3 Data Mapping and Conversion

Map external data to local format:

// Example: Apple Health steps mapping
function mapAppleHealthSteps(appleRecord) {
  return {
    date: formatDateTime(appleRecord.startDate),
    steps: parseInt(appleRecord.value),
    source: "Apple Health",
    device: appleRecord.sourceName
  };
}

// Save to local file
saveToLocalFile("data/fitness/activities.json", mappedData);

2.4 Data Validation

function validateImportedData(data) {
  // Check required fields
  // Validate data types
  // Check data ranges
  // Ensure correct time format

  return {
    valid: false,
    errors: ["Validation is not implemented; do not import until schema/unit/date checks pass"],
    warnings: []
  };
}

2.5 Generate Import Report

const importReport = {
  source: dataSource,
  import_date: new Date().toISOString(),
  records_imported: {
    steps: 1234,
    weight: 30,
    heart_rate: 1200,
    sleep: 90
  },
  date_range: {
    start: "2025-01-01",
    end: "2025-01-22"
  },
  validation: validationResults
};

Step 3: Knowledge Base Query Workflow

If user wants to query knowledge base:

3.1 Identify Query Topic

const topic = identifyTopic(userInput);
// Possible returns: "nutrition", "fitness", "sleep", "mental-health", "chronic-disease", "hypertension", "diabetes", etc.

3.2 Search Relevant Articles

Find relevant articles from knowledge base index:

function searchKnowledgeBase(topic) {
  // Read knowledge base index
  const kbIndex = readFile('.claude/skills/wellally-tech/knowledge-base/index.md');

  // Find matching articles
  const articles = kbIndex.categories.filter(cat =>
    cat.tags.includes(topic) || cat.keywords.includes(topic)
  );

  return articles;
}

3.3 Return Article Links

const results = {
  topic: topic,
  articles: [
    {
      title: "Hypertension Monitoring and Management",
      url: "https://wellally.tech/knowledge-base/chronic-disease/hypertension-monitoring",
      category: "Chronic Disease Management",
      description: "Learn how to effectively monitor and manage blood pressure"
    },
    {
      title: "Blood Pressure Lowering Strategies",
      url: "https://wellally.tech/knowledge-base/chronic-disease/bp-lowering-strategies",
      category: "Chronic Disease Management",
      description: "Improve blood pressure levels through lifestyle changes"
    }
  ],
  total_found: 2
};

Step 4: Intelligent Recommendation Workflow

If user wants personalized recommendations:

4.1 Read User Health Data

// Read relevant health data
const profile = readFile('data/profile.json');
const bloodPressure = glob('data/blood-pressure/**/*.json');
const sleepRecords = glob('data/sleep/**/*.json');
const weightHistory = profile.weight_history || [];

4.2 Analyze Health Status

function analyzeHealthStatus(data, reviewThresholds) {
  const status = {
    concerns: [],
    good_patterns: []
  };

  // Analyze blood pressure
  if (Number.isFinite(data.blood_pressure?.systolic) &&
      Number.isFinite(data.blood_pressure?.diastolic) &&
      (data.blood_pressure.systolic >= reviewThresholds.systolic ||
       data.blood_pressure.diastolic >= reviewThresholds.diastolic)) {
    status.concerns.push({
      area: "blood_pressure",
      severity: "high",
      condition: "Blood pressure reading to review; not a diagnosis",
      value: { systolic: data.blood_pressure.systolic, diastolic: data.blood_pressure.diastolic }
    });
  }

  // Analyze sleep
  if (data.sleep?.average_duration < 6) {
    status.concerns.push({
      area: "sleep",
      severity: "medium",
      condition: "Sleep Deprivation",
      value: data.sleep.average_duration + " hours"
    });
  }

  // Analyze weight trend
  if (data.weight?.trend === "increasing") {
    status.concerns.push({
      area: "weight",
      severity: "medium",
      condition: "Weight Gain",
      value: data.weight.change + " kg"
    });
  }

  // Identify good patterns
  if (data.steps?.average > 8000) {
    status.good_patterns.push({
      area: "activity",
      description: "Daily average steps over 8000",
      value: data.steps.average
    });
  }

  return status;
}

4.3 Recommend Relevant Articles

function recommendArticles(healthStatus) {
  const recommendations = [];

  for (const concern of healthStatus.concerns) {
    const articles = findArticlesForCondition(concern.condition);
    recommendations.push({
      condition: concern.condition,
      severity: concern.severity,
      articles: articles
    });
  }

  return recommendations;
}

4.4 Generate Recommendation Report

const recommendationReport = {
  generated_at: new Date().toISOString(),
  health_status: healthStatus,
  recommendations: recommendations,
  total_articles: recommendations.reduce((sum, r) => sum + r.articles.length, 0)
};

Security & Privacy

Must Follow

• ❌ Do not transmit private health data without explicit authorization for the exact service and data scope; API sync is not an exemption
• ❌ Do not hardcode API credentials in code
• ❌ Do not share user access tokens
• ✅ All imported data stored locally only
• ✅ OAuth credentials encrypted storage
• ✅ Import only after explicit user authorization

Data Validation

• ✅ Validate imported data types and ranges
• ✅ Filter abnormal values (e.g., negative steps)
• ✅ Preserve data source information
• ✅ Handle timezone conversion

Error Handling

File Read Failure:

• Output "Unable to read file, please check file path and format"
• Provide correct file format examples
• Suggest re-exporting data

API Call Failure:

• Output "API call failed, please check network connection and credentials"
• Provide OAuth re-authentication guidance
• Fall back to CSV import method

Data Validation Failure:

• Output "Incorrect data format, skipped invalid records"
• Log number of skipped records
• Continue processing valid data

Usage Examples

Example 1: Import Apple Health Data

User: "Import fitness tracker data from Apple Health"

Output: Execute import workflow, generate import report

Example 2: Query Knowledge Base

User: "WellAlly platform articl

… (nội dung đầy đủ có trong gói tải về)