What problem does it solve? Manually scanning dozens of RSS feeds, newsletters, and subreddits every day is noisy and time-consuming. This Skill automates the entire pipeline: it fetches new items from your configured sources, filters them against your stated interests, deduplicates repeated stories, summarizes the keepers, and writes a single clean markdown digest into an Obsidian vault. ## Core Features & Use Cases - Multi-source fetching: Pulls candidates from RSS 2.0 feeds, the 1440 daily newsletter (parsed per-blurb with citations), and subreddit listings, with per-source state tracking so nothing resurfaces. - LLM-based classification and dedup: Classifies items against your editable interests.md, clusters cross-outlet coverage of the same story, and suppresses stories already covered on previous days. - Paywall-aware summarization: Detects paywalled articles, searches for the same story at a non-paywalled source, and summarizes via WebFetch with feed-description fallbacks. - Use Case: Each morning, say "fetch my newspaper" and receive a timestamped digest in your Obsidian vault containing only the stories matching your interests, plus an inspectable Skipped section for tuning. ## Quick Start Say "run paperboy" to fetch your configured news sources, filter them against your interests, and write today's digest into your Obsidian vault.
