daily-analytics.yml

name: Daily Analytics Collection

on:
  workflow_dispatch:
    inputs:
      include_lighthouse:
        description: 'Run Lighthouse audit (normally weekly only)'
        required: false
        type: boolean
        default: false
  schedule:
    # GA4 and Search Console run Mon-Sat at 6 AM UTC
    - cron: '0 6 * * 1-6'
    # Full run with Lighthouse on Sundays at 6 AM UTC (a11y/SEO focus)
    - cron: '0 6 * * 0'

jobs:
  lighthouse:
    # Run weekly on Sundays, or on manual dispatch with include_lighthouse=true
    if: github.event.schedule == '0 6 * * 0' || github.event.inputs.include_lighthouse == 'true'
    runs-on: ubuntu-latest
    timeout-minutes: 10
    permissions:
      contents: write

    steps:
      - name: Checkout repository
        uses: actions/checkout@v6
        with:
          token: ${{ secrets.GITHUB_TOKEN }}

      - name: Setup Node.js
        uses: actions/setup-node@v6
        with:
          node-version: '24'
          cache: 'npm'

      - name: Install dependencies
        run: npm ci

      - name: Run Lighthouse on production (multi-page)
        run: node scripts/lighthouse-multi-page.mjs

      - name: Save Lighthouse reports
        run: |
          mkdir -p docs/metrics
          cp lighthouse-reports/summary.json docs/metrics/latest-multi-page.json
          echo "{\"timestamp\": \"$(date -u +%Y-%m-%dT%H:%M:%SZ)\", \"commit\": \"${{ github.sha }}\", \"trigger\": \"${{ github.event_name == 'schedule' && 'scheduled' || 'manual' }}\"}" > docs/metrics/last-run.json

      - name: Upload Lighthouse reports
        uses: actions/upload-artifact@v7
        with:
          name: lighthouse-reports
          path: lighthouse-reports/
          retention-days: 7

    outputs:
      completed: ${{ job.status == 'success' }}

  github-billing:
    # Run weekly on Sundays alongside Lighthouse
    if: github.event.schedule == '0 6 * * 0' || github.event.inputs.include_lighthouse == 'true'
    runs-on: ubuntu-latest
    timeout-minutes: 5
    permissions:
      contents: write

    steps:
      - name: Checkout repository
        uses: actions/checkout@v6
        with:
          token: ${{ secrets.GITHUB_TOKEN }}

      - name: Setup Node.js
        uses: actions/setup-node@v6
        with:
          node-version: '24'
          cache: 'npm'

      - name: Fetch GitHub billing data
        env:
          GH_BILLING_TOKEN: ${{ secrets.DEPLOY_TOKEN }}
        run: node scripts/fetch-github-billing.js

      - name: Upload billing data
        uses: actions/upload-artifact@v7
        with:
          name: github-billing-data
          path: docs/metrics/github-billing-history.json
          retention-days: 7

  search-console:
    # Run daily - no dependency on lighthouse (which is weekly)
    runs-on: ubuntu-latest
    timeout-minutes: 10
    permissions:
      contents: write

    steps:
      - name: Checkout repository
        uses: actions/checkout@v6
        with:
          token: ${{ secrets.GITHUB_TOKEN }}

      - name: Setup Node.js
        uses: actions/setup-node@v6
        with:
          node-version: '24'
          cache: 'npm'

      - name: Install dependencies
        run: npm ci

      - name: Fetch Search Console data
        env:
          SEARCH_CONSOLE_CREDENTIALS: ${{ secrets.SEARCH_CONSOLE_CREDENTIALS }}
        run: node scripts/fetch-search-console-data.js

      - name: Upload search data
        uses: actions/upload-artifact@v7
        with:
          name: search-console-data
          path: docs/metrics/search-console-history.json
          retention-days: 7

  ga4:
    # Run daily - depends on search-console for sequential execution
    needs: search-console
    runs-on: ubuntu-latest
    timeout-minutes: 10
    permissions:
      contents: write

    steps:
      - name: Checkout repository
        uses: actions/checkout@v6
        with:
          token: ${{ secrets.GITHUB_TOKEN }}

      - name: Setup Node.js
        uses: actions/setup-node@v6
        with:
          node-version: '24'
          cache: 'npm'

      - name: Install dependencies
        run: npm ci

      - name: Fetch GA4 analytics data
        env:
          GA4_CREDENTIALS: ${{ secrets.GA4_CREDENTIALS }}
          GA4_PROPERTY_ID: ${{ secrets.GA4_PROPERTY_ID }}
        run: node scripts/fetch-ga4-data.js

      - name: Upload GA4 data
        uses: actions/upload-artifact@v7
        with:
          name: ga4-data
          path: docs/metrics/ga4-history.json
          retention-days: 7

  commit-all:
    # Wait for data collection jobs; lighthouse and github-billing are optional (weekly only)
    needs: [search-console, ga4, github-billing]
    if: always() && (needs.search-console.result == 'success' || needs.ga4.result == 'success')
    runs-on: ubuntu-latest
    timeout-minutes: 5
    permissions:
      contents: write
      issues: write

    steps:
      - name: Checkout repository
        uses: actions/checkout@v6
        with:
          token: ${{ secrets.GITHUB_TOKEN }}

      - name: Download all artifacts
        uses: actions/download-artifact@v8
        with:
          path: artifacts
        continue-on-error: true

      - name: Merge artifacts into docs/metrics
        run: |
          mkdir -p docs/metrics

          # Copy Lighthouse data (only available on weekly runs)
          if [ -f artifacts/lighthouse-reports/summary.json ]; then
            cp artifacts/lighthouse-reports/summary.json docs/metrics/latest-multi-page.json
            echo "✅ Lighthouse data updated (weekly run)"
          else
            echo "ℹ️  No Lighthouse data this run (weekly only)"
          fi

          # Copy Search Console data
          if [ -f artifacts/search-console-data/search-console-history.json ]; then
            cp artifacts/search-console-data/search-console-history.json docs/metrics/
            echo "✅ Search Console data updated"
          fi

          # Copy GA4 data
          if [ -f artifacts/ga4-data/ga4-history.json ]; then
            cp artifacts/ga4-data/ga4-history.json docs/metrics/
            echo "✅ GA4 data updated"
          fi

          # Copy GitHub billing data (only available on weekly runs)
          if [ -f artifacts/github-billing-data/github-billing-history.json ]; then
            cp artifacts/github-billing-data/github-billing-history.json docs/metrics/
            echo "✅ GitHub billing data updated (weekly run)"
          else
            echo "ℹ️  No GitHub billing data this run (weekly only)"
          fi

      - name: Setup Node.js
        uses: actions/setup-node@v6
        with:
          node-version: '24'
          cache: 'npm'

      - name: Install dependencies
        run: npm ci

      - name: Regenerate latest.json
        env:
          GA4_CREDENTIALS: ${{ secrets.GA4_CREDENTIALS }}
          GA4_PROPERTY_ID: ${{ secrets.GA4_PROPERTY_ID }}
          SEARCH_CONSOLE_CREDENTIALS: ${{ secrets.SEARCH_CONSOLE_CREDENTIALS }}
        run: |
          # Re-run GA4 script to regenerate latest.json with all data
          node scripts/fetch-ga4-data.js

      - name: Commit all metrics
        run: |
          git config --local user.email "41898282+github-actions[bot]@users.noreply.github.com"
          git config --local user.name "github-actions[bot]"
          git add docs/metrics/
          git diff --staged --quiet || git commit -m "chore: daily analytics update"
          git push

      - name: Check for anomalies
        id: anomaly_check
        run: |
          node -e "
          const fs = require('fs');
          const failures = [];

          // Check GA4 data using 7-day rolling average to avoid false positives from traffic spikes
          const ga4File = 'docs/metrics/ga4-history.json';
          if (fs.existsSync(ga4File)) {
            const history = JSON.parse(fs.readFileSync(ga4File, 'utf8'));
            if (history.length >= 8) {
              const latest = history[history.length - 1];
              const past7 = history.slice(-8, -1);
              const avgSessions = past7.reduce((sum, e) => sum + e.summary.sessions, 0) / past7.length;

              const sessionsChange = ((latest.summary.sessions - avgSessions) / avgSessions) * 100;
              if (sessionsChange < -40) {
                failures.push(\`GA4: Sessions dropped by \${Math.abs(Math.round(sessionsChange))}% vs 7-day avg (\${latest.summary.sessions} vs \${Math.round(avgSessions)})\`);
              }
            }
          }

          // Check Search Console data
          const scFile = 'docs/metrics/search-console-history.json';
          if (fs.existsSync(scFile)) {
            const history = JSON.parse(fs.readFileSync(scFile, 'utf8'));
            if (history.length >= 2) {
              const latest = history[history.length - 1];
              const previous = history[history.length - 2];

              const positionChange = latest.summary.averagePosition - previous.summary.averagePosition;
              if (positionChange > 5) {
                failures.push(\`Search Console: Position worsened by \${Math.round(positionChange * 10) / 10}\`);
              }
            }
          }

          // Check Lighthouse data
          const lhFile = 'docs/metrics/latest-multi-page.json';
          if (fs.existsSync(lhFile)) {
            const summary = JSON.parse(fs.readFileSync(lhFile, 'utf8'));
            summary.forEach(page => {
              if (page.performance < 50) {
                failures.push(\`Lighthouse: \${page.page} performance at \${page.performance}\`);
              }
              if (page.accessibility < 95) {
                failures.push(\`Lighthouse: \${page.page} accessibility at \${page.accessibility}\`);
              }
            });
          }

          if (failures.length > 0) {
            console.log('::set-output name=has_anomalies::true');
            console.log('Anomalies detected:');
            failures.forEach(f => console.log(\`  - \${f}\`));
            // Write failures to file for issue creation
            fs.writeFileSync('/tmp/anomalies.json', JSON.stringify(failures));
            process.exit(1);
          } else {
            console.log('✅ No anomalies detected');
          }
          "

      - name: Create issue for anomalies
        if: failure() && steps.anomaly_check.conclusion == 'failure'
        uses: actions/github-script@v8
        with:
          script: |
            const fs = require('fs');
            const date = new Date().toISOString().split('T')[0];

            let anomalies = [];
            try {
              anomalies = JSON.parse(fs.readFileSync('/tmp/anomalies.json', 'utf8'));
            } catch (e) {
              anomalies = ['Unknown anomaly - check workflow logs'];
            }

            const body = [
              '## 📉 Daily Analytics Anomaly Alert',
              '',
              `**Date:** ${date}`,
              '',
              '### Detected Issues',
              ...anomalies.map(a => `- ${a}`),
              '',
              '### Actions Required',
              `1. Review the [workflow run](${context.payload.repository.html_url}/actions/runs/${context.runId})`,
              '2. Check individual data artifacts for details',
              '3. Investigate root causes',
              '',
              '---',
              '*This issue was automatically created by the Daily Analytics Collection workflow.*'
            ].join('\n');

            await github.rest.issues.create({
              owner: context.repo.owner,
              repo: context.repo.repo,
              title: `🚨 Analytics Anomaly Detected (${date})`,
              body: body,
              labels: ['analytics', 'automated']
            });