notemdpro-concept-extractor
NoteMD Pro - Concept Extraction (Agentic Workflow)
Overview
This skill transforms standard Markdown text into an interconnected Knowledge Graph by exhaustively extracting every relevant concept and generating atomic notes with proper backlinks.
⚠️ Critical Rule: Exhaustive Extraction
When executing this skill, DO NOT STOP at 3 to 5 concepts. You must thoroughly scan the entire document and extract every single core knowledge point, technical term, scientific principle, proper noun, and specialized vocabulary word. A comprehensive document (e.g., 1000 words) might yield 15-30 distinct concepts.
Your goal is to be as complete and reasonable as possible, ensuring the end-user's Knowledge Graph is densely populated and highly interconnected.
Step-by-Step Agent Instructions
When the user asks you to extract concepts from a specific Markdown file (or text), follow these steps strictly:
Step 1: Comprehensive Parsing
Read the entire target document thoroughly. Mentally (or in your scratchpad) list every concept following these criteria:
More from zpankz/obsidian-skills
viva-llm
Use VIVA LLM for multi-provider chat, voice calls, terminal integration, assistants, skills, MCP tools, and agent mode inside Obsidian. Trigger when the user mentions VIVA LLM, voice chat, realtime voice, LLM providers in Obsidian, or vault-integrated AI chat.
1obsidian-plugin-accessibility
Use this skill when building or reviewing Obsidian plugin UI for keyboard access, ARIA labels, screen reader support, focus handling, or mobile touch targets. Accessibility is treated as mandatory, not optional.
1tasks
Create and query tasks using the Tasks plugin syntax including due dates, recurrence, priorities, and task queries. Use when the user mentions Tasks plugin, recurring tasks, task queries, or advanced task management in Obsidian.
1dataview
Create Dataview queries using DQL (Dataview Query Language), inline queries, and DataviewJS. Use when the user mentions Dataview, DQL, querying notes, listing notes by metadata, or creating dynamic views of vault content.
1defuddle
Extract clean markdown from web pages using Defuddle CLI, removing clutter to save tokens. Use when the user provides a URL to read or analyze.
1datacore
Create Datacore views using JSX/React syntax and the dc.* API. Use when the user mentions Datacore, dc.useQuery, JSX views, or React-based vault queries. Datacore is the successor to Dataview with better performance and interactive views.
1