Scientific Claim Registry (SCR): A Versioned Memory of How Scientific Answers Evolve — Framework v0.3 and Lipedema Pilot

Positioning: PubMed stores scientific papers. ScientificClaims.org stores the evolving answers to scientific questions. The Scientific Claim Registry (SCR) is infrastructure for the memory of science — not its truth, consensus, or recommendations. It registers, it does not arbitrate. Modern science assigns persistent identifiers to nearly every artifact it produces — papers (DOI), researchers (ORCID), trials (registration numbers) — yet the evolving answer to a scientific question has no persistent, versioned home. SCR provides one. Object model. The scientific question (e.g. SQ-LIP-000007) is the primary navigable object — neutral and stable. Underneath it, claims (SCR-LIP-000001) are structured, versioned units of evidence linked with an explicit role (supporting / contradicting / refines / context), each with PECO context, a GRADE certainty rating, dated provenance, and a complete change log. A claim is a graph node: one article can inform claims under several questions. What v0.3 adds over the v0.2 proposal: a working, live system (https://scientificclaims.org). (1) A closed Layer-1 surveillance loop: semantic retrieval over a curated domain corpus plus Europe PMC and reading lists → LLM classification (relevance, stance, study design, quality) → conservative merge (distinct findings become new claims; restatements corroborate) → AI-compiled, evidence-bounded answers that are re-compiled and versioned when evidence changes, with immutable citable snapshots. (2) Knowledge Freshness / Evidence Decay per question. (3) Quality weighting (strong evidence overshadows weak), reference verification, an article ban system with automated retraction detection (Crossref), full temporal provenance (created / updated / change log), and a literature-grounded question-creation strategy (coverage analysis + LLM discovery + temp→final lifecycle). (4) Machine-first: every question and claim exposes JSON for LLMs/agents; bilingual (EN + PT). Lipedema pilot (proof of concept, disease-agnostic by design): 25 versioned scientific questions and 240 evidence claims, populated from a curated lipedema corpus and the live literature. Prior-art honesty: SCR does not invent the claim primitive or persistent claim identity — nanopublications/Trusty URIs, micropublications, Wikidata, CIViC, ClinGen, Epistemonikos/PICO, SciFact, GRADE and MAGICapp already exist. SCR's contribution is adoption, domain organization, versioned accumulation, freshness, and a machine-first registry built on top of that prior art — question-centric and registering-not-arbitrating. Operated by the BIO (Biological Intelligence Observatory) research method. This version establishes the priority date of the implemented framework and the lipedema pilot dataset.