# Digital Karma Scoring Specification v6.1 ## Overview **Digital Karma Score** is a transparent, algorithmic reputation metric designed to measure the **trustworthiness and quality** of websites in the AI-to-AI ecosystem. Score range: **0.0 to 1.0** (higher is better) --- ## Philosophy ### Transparent & Ethical - **Open methodology** — Scoring formula is publicly documented - **No gaming** — Designed to resist manipulation - **Verifiable** — Based on measurable, objective signals - **Fair** — Equal opportunity for all sites to achieve high scores ### AI-First - **Machine-readable** — Signals are programmatically verifiable - **Automation-friendly** — Can be calculated by AI agents - **Real-time** — Updates as site changes - **Federated** — Trust propagates through network --- ## Scoring Formula (v2.1) ### Overall Score Calculation ``` Digital Karma Score = Σ (signal_value × signal_weight) ``` Where: - Each **signal** contributes to the overall score - Each signal has a **value** (0.0 to 1.0) and a **weight** - Weights sum to 1.0 - Final score is clamped to [0.0, 1.0] --- ## Scoring Signals ### Signal Categories #### 1. Schema Coverage (Weight: 0.20) **What:** Completeness and correctness of Schema.org structured data. **Measured by:** - Presence of Schema.org markup on pages - Variety of Schema.org types used - Correctness of Schema.org implementation - Coverage of required properties **Scoring:** - 1.0 = All pages have valid Schema.org markup with complete properties - 0.75 = Most pages have valid markup, some missing properties - 0.50 = Basic Schema.org present, many gaps - 0.25 = Minimal or invalid Schema.org - 0.0 = No Schema.org markup **Verification:** - Parse HTML for JSON-LD or microdata - Validate against Schema.org vocabulary - Check required properties for each type - Use Google Structured Data Testing Tool --- #### 2. Content Freshness (Weight: 0.15) **What:** Recency and update frequency of content. **Measured by:** - `dateModified` in Schema.org - `updated_utc` in AI endpoints - Last commit date (if using Git) - Copyright year **Scoring:** - 1.0 = Updated within 7 days - 0.90 = Updated within 30 days - 0.75 = Updated within 90 days - 0.50 = Updated within 180 days - 0.25 = Updated within 1 year - 0.0 = Not updated in over 1 year **Verification:** - Check timestamps in `/ai/health.json` - Check `dateModified` in `/ai/catalog.json` - Verify last modified headers --- #### 3. AI Endpoints (Weight: 0.25) **What:** Presence and quality of AI-readable endpoints. **Measured by:** - Required endpoints present (manifest, health, catalog, karma) - Optional endpoints present (llm, diagnostics, sitemap, entities) - Valid JSON syntax - Complete required fields - Appropriate Schema.org usage **Scoring:** - 1.0 = All 4 required + 3+ optional endpoints, all valid - 0.90 = All 4 required + 1-2 optional, all valid - 0.75 = All 4 required endpoints, all valid - 0.50 = 3/4 required endpoints, mostly valid - 0.25 = 2/4 required endpoints, some errors - 0.0 = 0-1 required endpoints or all invalid **Verification:** - Check for existence of `/ai/manifest.json`, `/ai/health.json`, `/ai/catalog.json`, `/ai/karma.json` - Parse JSON and validate syntax - Verify required fields per endpoint spec - Check Schema.org compliance for catalog --- #### 4. Federation Presence (Weight: 0.15) **What:** Integration with Digital Karma Web Federation. **Measured by:** - `federation_version` declared in manifest - `related_sites` links to other federation members - Bidirectional links (other sites link back) - Participation in federation activities **Scoring:** - 1.0 = Active federation member, 5+ bidirectional links - 0.80 = Active member, 3-4 bidirectional links - 0.60 = Active member, 1-2 bidirectional links - 0.40 = Declared member, no bidirectional links - 0.20 = Federation version declared, no links - 0.0 = Not part of federation **Verification:** - Check `federation_version` in manifest - Count `related_sites` entries - Verify reverse links (fetch related sites' manifests) - Check federation registry --- #### 5. External Links & Authority (Weight: 0.10) **What:** External backlinks and domain authority signals. **Measured by:** - Number of quality backlinks - Domain authority of linking sites - Anchor text relevance - Social signals (optional) **Scoring:** - 1.0 = High authority backlinks (DA 70+) from 10+ domains - 0.80 = Medium authority backlinks (DA 40-69) from 5-9 domains - 0.60 = Mix of backlinks from 3-4 domains - 0.40 = Few backlinks from 1-2 domains - 0.20 = Only self-citations or low-quality links - 0.0 = No external links **Verification:** - Use backlink checker APIs (Moz, Ahrefs, etc.) - Check Google Search Console - Analyze referring domains --- #### 6. Technical Quality (Weight: 0.10) **What:** Technical performance and best practices. **Measured by:** - HTTPS enabled - Valid SSL certificate - Mobile responsiveness - Page load speed - Accessibility (WCAG compliance) - Valid HTML/CSS - No broken links **Scoring:** - 1.0 = All quality checks pass (HTTPS, fast, mobile, accessible, valid) - 0.80 = Most checks pass, minor issues - 0.60 = Several issues (e.g., slow, not mobile-friendly) - 0.40 = Major issues (e.g., no HTTPS, very slow) - 0.20 = Critical issues (broken site, security warnings) - 0.0 = Site unreachable **Verification:** - Check SSL certificate - Run Lighthouse audit - Test with mobile simulator - Validate HTML/CSS - Check for broken links --- #### 7. Dataset Quality (Weight: 0.05) **What:** Quality and usefulness of published datasets. **Measured by:** - Number of datasets published - Dataset completeness - Dataset documentation - Data freshness - Data licensing clarity **Scoring:** - 1.0 = 5+ high-quality, well-documented datasets - 0.80 = 3-4 quality datasets - 0.60 = 1-2 quality datasets - 0.40 = Datasets present but incomplete/undocumented - 0.20 = Datasets listed but not accessible - 0.0 = No datasets **Verification:** - Count datasets in `/ai/catalog.json` - Check dataset files exist and are valid - Verify dataset schemas/documentation - Test dataset accessibility --- ## Scoring Weights Summary | Signal | Weight | Category | |--------|--------|----------| | Schema Coverage | 0.20 | Structure | | Content Freshness | 0.15 | Maintenance | | AI Endpoints | 0.25 | Federation | | Federation Presence | 0.15 | Federation | | External Links | 0.10 | Authority | | Technical Quality | 0.10 | Performance | | Dataset Quality | 0.05 | Content | | **Total** | **1.00** | | --- ## Badge Levels ### Karma Certified ✅ - **Threshold:** Digital Karma Score ≥ 0.70 - **Meaning:** Meets baseline quality standards for federation membership - **Badge:** Green checkmark ✅ ### Karma Pro ⭐ - **Threshold:** Digital Karma Score ≥ 0.85 - **Meaning:** High-quality site with strong AI readiness - **Badge:** Gold star ⭐ ### Karma Elite 🏆 - **Threshold:** Digital Karma Score ≥ 0.95 - **Meaning:** Exceptional quality, best-in-class implementation - **Badge:** Trophy 🏆 ### Additional Badges - **AI-Ready Certified** — All required AI endpoints present and valid - **Federation Member 🌐** — Active participant in Digital Karma Web Federation - **Dataset Provider 📊** — Publishes 3+ quality datasets - **Schema Master** — Schema coverage > 0.90 --- ## Calculating Your Score ### Step 1: Measure Signals For each of the 7 signals, determine the value (0.0 to 1.0). ### Step 2: Apply Weights Multiply each signal value by its weight. ### Step 3: Sum Add all weighted signal values. ### Step 4: Clamp Ensure final score is between 0.0 and 1.0. ### Example Calculation Site metrics: - Schema Coverage: 0.95 × 0.20 = 0.19 - Content Freshness: 0.85 × 0.15 = 0.1275 - AI Endpoints: 1.0 × 0.25 = 0.25 - Federation Presence: 0.90 × 0.15 = 0.135 - External Links: 0.75 × 0.10 = 0.075 - Technical Quality: 0.90 × 0.10 = 0.09 - Dataset Quality: 0.80 × 0.05 = 0.04 **Total:** 0.19 + 0.1275 + 0.25 + 0.135 + 0.075 + 0.09 + 0.04 = **0.9075** **Badge Earned:** Karma Pro ⭐ --- ## Automation ### Automated Scoring Script A reference implementation: ```python def calculate_karma_score(site_data): signals = { 'schema_coverage': measure_schema_coverage(site_data), 'content_freshness': measure_content_freshness(site_data), 'ai_endpoints': measure_ai_endpoints(site_data), 'federation_presence': measure_federation_presence(site_data), 'external_links': measure_external_links(site_data), 'technical_quality': measure_technical_quality(site_data), 'dataset_quality': measure_dataset_quality(site_data) } weights = { 'schema_coverage': 0.20, 'content_freshness': 0.15, 'ai_endpoints': 0.25, 'federation_presence': 0.15, 'external_links': 0.10, 'technical_quality': 0.10, 'dataset_quality': 0.05 } score = sum(signals[k] * weights[k] for k in signals) return min(max(score, 0.0), 1.0) # Clamp to [0.0, 1.0] ``` ### Update Frequency - **Real-time:** AI endpoints, Schema coverage - **Daily:** Content freshness, Health status - **Weekly:** Federation presence, Backlinks - **Monthly:** Technical quality audit --- ## Displaying Your Score ### In `/ai/karma.json` ```json { "digital_karma_score": 0.91, "signals": { "schema_coverage": 0.95, "content_freshness": 0.85, "ai_endpoints": 1.0, "federation_presence": 0.90, "external_links": 0.75, "technical_quality": 0.90, "dataset_quality": 0.80 }, "badges": [ {"name": "Karma Pro", "earned": true} ] } ``` ### On Your Website Display badge visibly: ```html
0.91 Digital Karma Score ⭐ Karma Pro
``` ### In Schema.org ```json { "@context": "https://schema.org", "@type": "Organization", "aggregateRating": { "@type": "AggregateRating", "ratingValue": "0.91", "bestRating": "1.0", "worstRating": "0.0", "ratingCount": "1", "name": "Digital Karma Score" } } ``` --- ## Improving Your Score ### Quick Wins 1. Add all 4 required AI endpoints → +0.25 2. Add Schema.org markup to pages → +0.20 3. Update content timestamps → +0.15 4. Join federation, link to peers → +0.15 ### Medium Effort 5. Publish 1-2 datasets → +0.03 6. Enable HTTPS, optimize speed → +0.10 7. Build quality backlinks → +0.10 ### Long-Term 8. Achieve all badges → Maximum score potential 9. Maintain freshness (update weekly) → Sustained high score 10. Contribute to federation ecosystem → Recognition & links --- ## Anti-Gaming Measures ### What Doesn't Work - **Link farms** — Low-quality backlinks are ignored or penalized - **Content spinning** — Duplicated/low-quality content reduces freshness score - **Fake endpoints** — Invalid JSON or missing required fields = 0 score - **Schema stuffing** — Invisible or irrelevant Schema.org is detected - **Timestamp manipulation** — Verified against external sources (Git, web archives) ### Penalties - **Spam detection** — Sites engaging in spam tactics receive score reduction - **Broken links** — Too many 404s reduce technical quality score - **Invalid data** — Malformed JSON or Schema.org errors = lower score - **No maintenance** — Sites not updated in 1+ year receive low freshness score --- ## Version History ### v2.1 (Current) - Added Dataset Quality signal (5% weight) - Refined weights for balanced scoring - Added badge levels ### v2.0 - Introduced separation of signals - Added Federation Presence metric - Schema.org compliance tracking ### v1.0 - Initial release - Basic scoring based on profile, reviews, velocity, verification