Keep every finding traceable
The library checks source identity and preserves the original artifact hash alongside extracted findings. Published measurements retain their units, denominator, material, conditions, method, size context, and location in the paper.
Keep the boundaries visible
A missing value is omitted. A chemical concentration is not a particle count. A laboratory release experiment is not a measured human dose or a clinical outcome. Results from unlike studies remain separate.
Coverage
The collection follows the supplied category library. Every supplied PDF was inventoried, hash-checked and screened for topic relevance. Selected original papers were examined in their methods and results; reviews, misfiled papers and candidates not deeply reviewed are recorded separately. Category pages are selected literature syntheses, not exhaustive systematic reviews. A folder containing only a review cannot supply an invented product measurement.