What problem does it solve? Capturing ideas, papers, and research findings into a structured personal knowledge base requires consistent formatting, classification, and validation. This Skill turns conversational input or pasted URLs into validated knowledge card files without manual YAML editing. ## Core Features & Use Cases - Structured Card Creation: Drafts title, summary, epistemic type, confidence, source tier, domains, and content categories, then writes the card through the construct knowledge card create CLI. - Schema Enforcement: Delegates file operations to the Python CLI, which validates domains, categories, confidence ranges, and epistemic type enums, returning structured JSON errors on failure. - Connection Suggestions: Scans existing cards for related, supporting, or contradicting entries and optionally adds typed connections via construct knowledge connection add. - Use Case: You paste a link to a paper on successor representations; the Skill extracts key findings, classifies the card, creates it in cards/, and proposes a connection to an existing spatial reasoning card. ## Quick Start Tell the agent "add a card for this paper" and paste the URL or describe the finding you want captured.
