Digital Karma Scoring Methodology

Understanding the transparent, automated trust and quality scoring system for AI websites.

The Seven Signals

  1. Schema Coverage (20%) - Quality and completeness of Schema.org markup
  2. Content Freshness (15%) - Update frequency and recency
  3. AI Endpoints (25%) - Presence and completeness of required machine-readable files
  4. Constellation Presence (15%) - Constellation participation and bidirectional alignments
  5. External Links (10%) - Quality and relevance of outbound references
  6. Technical Quality (10%) - Performance, security, accessibility
  7. Dataset Quality (5%) - Availability of structured datasets

Badge Levels

  • Constellation Member - All 6 required artifacts present plus root robots.txt and sitemap.xml
  • Karma Certified - Score >= 0.70
  • Karma Pro - Score >= 0.85
  • Karma Elite - Score >= 0.95

Calculating Your Score

Run python ai/calculate_karma.py against your site to compute your score automatically. The script reads your AI endpoints and applies the weighted formula. Results are written directly to /ai/karma.json.

See the Standards page for the complete formula and signal breakdown.

Constellation Audit

Check a live site against what you just read

Use the constellation audit to compare a public implementation against the current Digital Karma rules and export a starter upgrade package.

Public URLs only. Instant results. Exportable upgrade package.

Audit a Site Back to Standards
← Back to Knowledge Base