Machine-readable supplement for "SAFE-T: A Survey and Framework for Trustworthy Healthcare AI Agents in the Internet of Medical Things"

Machine-readable supplement for the SAFE-T survey. Version v3 accompanies the first revision of the manuscript. It contains the archived database search snapshot of 4 October 2026 (10,740 records) with the disposition of every record; the reconciliation of the 256 cited works against it; and the evidence-coverage ledger of 232 anchors (226 healthcare-agent studies and six adjacent references) across 35 coded cells, with the coding rules, the census of cited works, the title-and-abstract screen of the snapshot, the four dimension-targeted searches, and their full-text checks. It also holds the four codings of every anchor, the adjudication, the agreement and sensitivity analyses, and the field codings of evidence maturity level, data setting, and fault tests. The 11-anchor ledger of the original submission is kept for comparison. README.md inside the archive describes every file; scripts/evidence/verify_bundle.py re-derives the ledgers from the snapshot and reports any discrepancy.