# Snapshot method — v2

Audit date: 2026-10-26. Source of record: COMPLETE catalog, including a library copy of `complete-archive.csv` (145 catalog rows). Seven entries retain stable snapshot IDs derived from one-based CSV row positions. Five extra records are explicitly *dossier locators* provided by the brief; their page content was not independently verified by the snapshot. No reconstructed creation dates, owners, completeness percentages or project outcomes are asserted.

The original catalog marks the seven selected entries `Listed · not uptime-verified`; displaying their URL as a functioning public website would misstate the evidence. These remain visible as disabled text, rather than clickable links. Dossier anchors are presented only as archive coordinates. Curatorial relationships and questions are newly written commentary, not historical evidence. The terms `private`, `dormant`, `proposed`, `abandoned` and `failed` are *visitor hypotheses*, not imported verified states.

The site bundles this immutable snapshot and does not scrape or sync the source at runtime. Future updates require human inspection and a versioned audit. Local visitor decisions are not edits to the archive.