What problem does it solve? Generic recommendation feeds tend to reinforce existing interests, creating a filter bubble. This Skill builds a small, surprising Wikipedia reading list that balances what you already like with adjacent subjects and genuine wildcards, turning idle curiosity into a structured learning path. ## Core Features & Use Cases - Interest Inference: Reads the current conversation to infer your interests, falling back to a broad mix when context is thin. - Verified Recommendations: Searches the live web to confirm every recommended article exists and uses its canonical URL. - Balanced Curation: Selects 5-8 articles split between known interests, bridging subjects, and at least one unexpected wildcard, each with a short hook and a closing synthesis question. - Use Case: Ask for a daily Wikipedia digest and receive a scannable list of linked articles with one- or two-sentence explanations of why each is worth reading and how it connects to your interests. ## Quick Start Ask the assistant to use discover-wikipedia to create today's Wikipedia reading list based on your interests.