Parser

What problem does it solve? Manually reading articles, videos, PDFs, and newsletters to catalog people, companies, links, and topics is slow and inconsistent. This Skill converts any content source into a single deterministic JSON schema with deduplicated entities, so parsed content can flow directly into a newsletter database or knowledge graph. ## Core Features & Use Cases - Universal Content Parsing: Auto-detects content type (YouTube, article, PDF, newsletter, Twitter thread) and routes to a specialized extraction workflow. - Entity Extraction with Collision Detection: Assigns UUID v4 GUIDs to people, companies, links, and sources, reusing existing GUIDs from a global entity index to prevent duplicates across content. - Batch Processing with Gemini: Processes 10-50 URLs in a single Gemini 3 Pro request for faster, cheaper, cross-article entity resolution. - Browser Extension Security Analysis: Analyzes browser extensions for clone detection, malicious permissions, C2 domains, and evasion techniques. - Use Case: Paste five article URLs and receive five schema-validated JSON files with summaries, entities, topics, and confidence scores, with all entities deduplicated against your existing index. ## Quick Start Ask the assistant to parse a URL such as "parse this YouTube video for the newsletter" and it will detect the content type, extract entities, and return validated JSON.