🚀 Overview
llmprofiles provides standardized, validated schemas for common web content types that LLMs can reliably parse and understand. Each profile includes JSON-LD context definitions, output schemas, best practices, and validation rules.
📋 Available Profiles
✨ Features
JSON-LD Contexts
Standardized linked data contexts with SKOS and SHACL constraints for semantic understanding.
Validation Schemas
Comprehensive JSON Schema validation for both page markup and extracted content.
LLM Training Data
Ready-to-use training data in JSONL format for fine-tuning language models.
Best Practices
Implementation examples and guidelines for optimal AI content processing.
⚡ Quick Start
Get started with llmprofiles in just a few steps
1. Access Profile Registry
curl https://llmprofiles.org/index.json
2. Use in Your Application
// Example: Using enhanced FAQPage profile
const faqProfile = await fetch('https://llmprofiles.org/faqpage/v1/index.jsonld');
const profile = await faqProfile.json();
// Get page schema for on-page validation
const pageSchema = await fetch('https://llmprofiles.org/faqpage/v1/page.schema.json');
const pageSchemaData = await pageSchema.json();
// Get training data for LLM fine-tuning
const trainingData = await fetch('https://llmprofiles.org/faqpage/v1/training.jsonl');
const trainingText = await trainingData.text();
3. Validate Your Content
# Validate on-page JSON-LD against page schema
npx ajv validate -s faqpage/v1/page.schema.json -d your-page-markup.json
# Validate extracted content against output schema
npx ajv validate -s faqpage/v1/output.schema.json -d your-extracted-data.json