What problem does it solve? Reading dense research papers raises questions that the paper alone often cannot answer, and ad-hoc answers lose provenance, contradict earlier notes, or blur the line between what the paper states and what is inferred. This Skill acts as a persistent answerer that grounds every response in traceable evidence. ## Core Features & Use Cases - Layered Evidence Answers: Separates paper evidence (source blocks, equations, figures, tables), external evidence (primary papers, official docs, forums), inference, and uncertainty in every answer note. - Persistent Question Ledger: Maintains stable question IDs, parent/related links, follow-up threads, and statuses such as unanswerable, open-research, and later_section across reading sessions. - External Research Integration: Researches beyond the paper using primary literature, official implementations, documentation, and forums, logging searches and sources in research notes. - Use Case: While reading a machine learning paper, you ask why a baseline underperforms. The Skill cites the paper's ablation table, checks the official implementation repository, labels forum-sourced interpretation as such, and saves a linked answer note in the workspace. ## Quick Start Use the research-paper-questions skill to answer my active questions about the current paper with cited evidence.
