What problem does it solve? Paywalled or restricted sources often get prematurely marked inaccessible, forcing investigators to downgrade confidence or drop evidence. This Skill enforces a structured, legal access hierarchy so every retrieval avenue is attempted before a source is written off. ## Core Features & Use Cases - Ordered Access Decision Tree: Steps through free-version search, Unpaywall, CORE, Semantic Scholar, OpenAlex, web archives, Google Scholar, reader mode, and author contact until full text is found. - Access Method Taxonomy: Records an access_method value (e.g., open_access, archive_copy, abstract_only) in findings.json and fact-check.json with a confidence-cap table tying each method to a maximum confidence level. - Safety and Data Minimization: Validates DOIs, URLs, and timestamps via spotlight_safe.py before shell calls, and restricts what data external services like Unpaywall or CORE may receive. - Use Case: A fact-checker hits a hard paywall on a key academic paper; the Skill walks them through Unpaywall, CORE, and Wayback Machine snapshots, retrieves a pre-paywall archived copy, and logs access_method: archive_copy with the snapshot date. ## Quick Start Use the content-access skill to retrieve the full text of this paywalled paper and record how it was accessed in findings.json.
