What problem does it solve? Reading and evaluating academic papers rigorously is slow and error-prone: claims get conflated with critique, metadata goes unverified, and long documents overflow context. This Skill provides a disciplined workflow for ingesting, summarizing, and critiquing a single research paper while tracking uncertainty and provenance. ## Core Features & Use Cases - Multi-source ingestion: Accept papers as PDFs, URLs, arXiv links, DOIs, or pasted text, preserving source provenance and access dates. - Structured extraction: Pull title, authors, venue, abstract, problem, method, evidence, and limitations, then separate the paper's claims from your own critique. - Bounded long-document reading: Read large documents in sized windows or parallel chunks with overlap handling, boundary-partial marking, and coverage-gap reporting so interrupted reads lose nothing. - Use Case: Before writing a design doc that depends on a paper's method, ingest the actual source rather than an abstract, extract its assumptions and limitations, and record a verified note with follow-up papers to read. ## Quick Start Read this arXiv paper, summarize its contribution and limitations, and critique whether its evidence supports my planned approach.
