@yottagraph-app/data-model-skill

Data model skill documentation for AI agents - entity types, properties, and schemas from Lovelace fetch sources.

Overview

This package provides documentation about the data model used by Lovelace's fetch sources. Each source (EDGAR, FRED, FDIC, etc.) contributes entity types, properties, and relationships to the knowledge graph. This skill helps AI agents understand what data is available and how it's structured.

Installation

This package is published to public npm:

npm install @yottagraph-app/data-model-skill

Usage

const dataModel = require('@yottagraph-app/data-model-skill');

// Get the main skill file path
console.log(dataModel.skillPath);

// List available sources
console.log(dataModel.sources); // ['edgar', 'fdic', 'fred', ...]

// Get a specific source file
const edgarSchema = dataModel.getSourceFile('edgar', 'schema.yaml');
const edgarDict = dataModel.getSourceFile('edgar', 'DATA_DICTIONARY.md');

Contents

The skill/ directory contains:

• SKILL.md - Main skill entry point with overview and navigation
• overview.md - Detailed concepts and per-source summary
• <source>/ - Per-source directories containing:
  • DATA_DICTIONARY.md - Human-readable documentation of entity types, properties, and relationships
  • schema.yaml - Machine-readable schema definition

Sources

| Source | Description | |--------|-------------| | edgar | SEC EDGAR filings (10-K, 10-Q, 8-K, ownership forms, etc.) | | fdic | FDIC BankFind Suite (insured institutions, failures) | | fred | Federal Reserve Economic Data (GDP, employment, rates) | | newsdata | News articles and press releases | | polymarket | Polymarket prediction markets | | sanctions | OpenSanctions (OFAC, EU, UN sanctions lists) |

Development

To sync the skill files from the source repository:

npm run sync

This copies DATA_DICTIONARY.md and schema.yaml files from moongoose/fetch/fetchtypes/ into the skill/ directory.

Exclusion Config

Edit sync-config.json to exclude sources:

{
  "exclude": ["gdelt", "wikipedia"]
}

License

MIT