note: "Every upstream source this corpus consumes. Human-approved via PR BEFORE any\ningestion. GENERATED\ \ by src/enumerate_audits.py — do not hand-edit; re-run it.\n\nSource: the SharePoint list behind sos.oregon.gov/audits/pages/state-audits.aspx,\n\ which renders nothing server-side. The list is captioned by the division itself as\n'State Audits and\ \ Reviews 2020 to present', which is why this corpus floors at 2020.\n\nTraps recorded so a future re-enumeration\ \ does not fall into them:\n1. The search page's HTML carries a fresh __VIEWSTATE per request, so its\ \ bytes are\n never stable. Compare the extracted RECORD SET, never the page.\n2. hrefs inside Additional_x0020_Documents\ \ are entity-escaped (`https://`).\n Unescape them or every URL 404s in a way that looks exactly\ \ like link rot.\n3. Report numbers are NOT dense. A gap can be a withdrawal under GAGAS 3.34 — one\n\ \ really is — so every gap is listed below with a reason rather than skipped.\n4. `format: html` is\ \ deliberate and is NOT a mistake: these URLs serve an HTML\n viewer with the PDF base64-encoded inside\ \ a