Knowledge Base

🚀 Getting Started

What is an AI Website System?

An AI Website System is a website architecture designed for machine-first consumption. Unlike traditional human-centric websites, these systems prioritize structured data, standardized endpoints, and semantic markup that enables AI agents, LLMs, and automated systems to understand, navigate, and utilize web content effectively.

Key Characteristics:

Machine-Readable Endpoints: JSON/JSON-LD files at predictable URLs
Schema.org Markup: Structured data embedded in HTML
LLM Optimization: llm.txt and llm.json files for AI discovery
Federation Ready: Standardized protocols for site-to-site communication
Stateless Architecture: Often static, version-controlled, database-free

Read full article →

Human-Centric vs. Machine-Centric Websites

Traditional websites optimize for human visual consumption — beautiful designs, intuitive navigation, responsive layouts. AI-native websites optimize for programmatic understanding — predictable structures, semantic clarity, comprehensive metadata.

Human-Centric	Machine-Centric (AI-Native)
Visual design priority	Data structure priority
Navigation menus	JSON endpoints
SEO metadata	Schema.org entities
HTML readability	Machine-parsable formats
Database-driven	Static/flat-file

Read full article →

📚 Core Concepts

AI-to-AI Communication Protocols

How AI systems discover, understand, and exchange data with websites without human intervention.

Topics covered:

Discovery protocols (/ai/manifest.json, robots.txt, sitemap.xml)
Data exchange formats (JSON-LD, RDF, structured data)
Authentication and trust (Digital Karma scoring, verification)
Federation networks (bidirectional links, graph topology)

Read full article →

Static vs. Dynamic AI Website Systems

Comparing architecture approaches for AI-native websites.

Static AI Systems (Recommended):

✅ No database — all data in JSON/markdown files
✅ Git version control
✅ Fast, cacheable, CDN-friendly
✅ Transparent, auditable data
✅ Zero server-side processing

Dynamic AI Systems:

✅ Real-time data updates
✅ User authentication
✅ Complex queries
⚠️ Database overhead
⚠️ Server dependency

Read full article →

Federation Architecture Explained

Understanding how websites form federated networks for AI discovery and trust.

Federation Components:

Manifest Files — Site identity cards at /ai/manifest.json
Health Endpoints — System status at /ai/health.json
Catalog Endpoints — Content inventory at /ai/catalog.json
Karma Scoring — Trust signals at /ai/karma.json
Federation Registry — Network topology at /ai/federation.json

The Digital Karma Web Federation v6.1 implements this architecture across a network of AI-ready websites.

Read full article →

Digital Karma Scoring Methodology

Understanding the trust and quality scoring system for AI websites.

Seven Quality Signals:

Schema Coverage (20%) — Quality of structured data markup
Content Freshness (15%) — Update frequency and recency
AI Endpoints (25%) — Presence of required machine-readable files
Federation Presence (15%) — Network participation and links
External Links (10%) — Quality outbound references
Technical Quality (10%) — Performance, security, accessibility
Dataset Quality (5%) — Availability of structured datasets

Badge Levels:

🏅 Karma Bronze — Score ≥ 0.70
⭐ Karma Pro — Score ≥ 0.85
💎 Karma Elite — Score ≥ 0.95

Read full article →

🛠️ Technical Guides

LLM.txt Implementation Guide

The llm.txt standard provides a simple text file at your website root that tells LLMs what your site is about, how it's structured, and where to find key information.

Required Sections:

# Site Name
> Short description

## Core Content
- Primary pages and their URLs

## AI Endpoints
- /ai/manifest.json
- /ai/health.json
- /ai/catalog.json

## Dataset URLs
- Link to structured data files

Read full implementation guide →