What problem does it solve? Saving articles, videos, PDFs, and posts into a personal knowledge base usually produces isolated summaries that are never linked to what you already know. This Skill files one external source into a wiki as a verbatim raw copy plus a synthesized entry that is cross-linked against existing entries, so new material integrates into the knowledge graph instead of sitting alone. ## Core Features & Use Cases - Multi-source fetching: Handles ordinary web pages, YouTube transcripts, PDFs, X/Twitter posts, Medium articles, and JavaScript-rendered pages, each with a dedicated fetcher that saves the verbatim original under raw/. - Gated filing with quality checks: A filing script refuses entries missing a TL;DR, a Related section with two or more wiki links, or valid frontmatter, and verifies every blockquote against the saved raw source. - Duplicate detection and revision tracking: Normalized URL comparison skips sources already in the wiki, while --revises files a later snapshot of an evolving source. - Direct and staged modes: File immediately with backlinks added to related entries, or stage to _inbox/proposed/ for later promotion by /wiki-promote in batch workflows. - Use Case: You find a YouTube talk relevant to your research notebook. Run the skill with the URL; it fetches the transcript, searches your wiki for related entries, writes a synthesis with attributed quotes, links it to at least two existing entries, and files it under the right taxonomy folder. ## Quick Start Ask the AI to add this article to the wiki by running /wiki-update followed by the URL of the source you want to file.
