Fresh for Which Claim? Separating evidence age, source lineage and decision scope in supplier research

A recently opened supplier profile can contain evidence of very different ages. Its address may have been checked today, a certification months earlier, and a delivery statement copied from an undated source. Treating the profile as uniformly current obscures the question a buyer actually needs answered: which claim is supported for this decision, by which source, and at what time?This paper develops a claim-level evidence model separating collection, substantive verification, stated validity, observed status, source lineage and applicability. Three small synthetic modules examine temporal state, historical reconstruction and scope isolation. Fourteen temporal, thirteen history/lineage and ten scope assertions pass separately; these are specification checks, not supplier accuracy measurements. Two scope checks reproduce a defect in the exploratory history routine, while eight exercise a partitioning wrapper. An author-reported shipping experience illustrates why a duration also needs a defined starting event and destination scope. The contribution is a proposed reporting and review discipline, not a universal freshness score, live supplier verification service or empirically optimized refresh policy.This is an archival mirror of the unchanged working paper v1.0, originally deposited on Zenodo on 3 November 2026. Not peer reviewed.Canonical citation: https://doi.org/10.5281/zenodo.23899662Manuscript and reproducibility materials: https://zenodo.org/records/23899662Mohammad Maani established the research direction, supervised the work, supplied methodological decisions and an illustrative shipping account, reviewed and accepted the scope-isolation interpretation and figure, and authorized publication. OpenAI ChatGPT/Codex assisted discovery, synthesis, drafting, research code and synthetic test design and agent execution; OpenAI image generation produced the conceptual artwork. Independent human code replication and source-by-source verification are not claimed. Full contribution and limitations disclosures are included in the manuscript.Competing interest: the author is Founder of QuickAIBuy and has a commercial interest in its development.Project: https://www.quickaibuy.comLicense: CC BY 4.0, subject to the manuscript's separate trademark, font and generated-artwork rights qualifications.