diff --git a/backend/scripts/import_notes_from_csv.py b/backend/scripts/import_notes_from_csv.py
new file mode 100644
index 0000000..b7689a2
--- /dev/null
+++ b/backend/scripts/import_notes_from_csv.py
@@ -0,0 +1,174 @@
+#!/usr/bin/env python3
+"""
+import_notes_from_csv.py
+------------------------
+Mass-import finding notes from a CSV file into the CVE dashboard database.
+
+CSV format (header row required, column names are case-insensitive):
+    ID,NOTES
+    12345,EXC-5754
+    67890,EXC-6001 - pending review
+
+Usage:
+    python3 import_notes_from_csv.py <csv_file> [--db <db_path>] [--dry-run]
+
+Options:
+    --db <path>   Path to cve_database.db  (default: ../cve_database.db)
+    --dry-run     Print what would change without touching the database
+"""
+
+import csv
+import sqlite3
+import sys
+import os
+import argparse
+from datetime import datetime, timezone
+
+NOTE_MAX_LEN = 255
+
+DEFAULT_DB = os.path.join(os.path.dirname(__file__), '..', 'cve_database.db')
+
+
+def parse_args():
+    p = argparse.ArgumentParser(description='Import finding notes from CSV into the dashboard DB.')
+    p.add_argument('csv_file', help='Path to the CSV file (must have ID and NOTES columns)')
+    p.add_argument('--db', default=DEFAULT_DB, help=f'Path to SQLite database (default: {DEFAULT_DB})')
+    p.add_argument('--dry-run', action='store_true', help='Preview changes without writing to DB')
+    return p.parse_args()
+
+
+def load_csv(path):
+    """Read CSV and return list of (finding_id, note) tuples."""
+    rows = []
+    with open(path, newline='', encoding='utf-8-sig') as f:
+        reader = csv.DictReader(f)
+        # Normalise header names to uppercase for case-insensitive matching
+        if reader.fieldnames is None:
+            print('ERROR: CSV file is empty or has no header row.')
+            sys.exit(1)
+
+        normalised = {k.strip().upper(): k for k in reader.fieldnames}
+        if 'ID' not in normalised or 'NOTES' not in normalised:
+            print(f'ERROR: CSV must have "ID" and "NOTES" columns.')
+            print(f'       Found columns: {list(reader.fieldnames)}')
+            sys.exit(1)
+
+        id_col    = normalised['ID']
+        notes_col = normalised['NOTES']
+
+        for i, row in enumerate(reader, start=2):  # start=2 because row 1 is the header
+            finding_id = row[id_col].strip()
+            note       = row[notes_col].strip()
+
+            if not finding_id:
+                print(f'  WARNING row {i}: empty ID — skipping')
+                continue
+
+            if len(note) > NOTE_MAX_LEN:
+                print(f'  WARNING row {i} ({finding_id}): note is {len(note)} chars, '
+                      f'truncating to {NOTE_MAX_LEN}')
+                note = note[:NOTE_MAX_LEN]
+
+            rows.append((finding_id, note))
+
+    return rows
+
+
+def run(args):
+    csv_path = os.path.abspath(args.csv_file)
+    db_path  = os.path.abspath(args.db)
+
+    # ------------------------------------------------------------------ checks
+    if not os.path.exists(csv_path):
+        print(f'ERROR: CSV file not found: {csv_path}')
+        sys.exit(1)
+
+    if not os.path.exists(db_path):
+        print(f'ERROR: Database not found: {db_path}')
+        sys.exit(1)
+
+    print(f'CSV : {csv_path}')
+    print(f'DB  : {db_path}')
+    if args.dry_run:
+        print('MODE: DRY RUN — no changes will be written\n')
+    else:
+        print()
+
+    # ----------------------------------------------------------------- load CSV
+    rows = load_csv(csv_path)
+    if not rows:
+        print('No valid rows found in CSV.')
+        sys.exit(0)
+
+    print(f'Loaded {len(rows)} row(s) from CSV.\n')
+
+    # ---------------------------------------------------------------- open DB
+    con = sqlite3.connect(db_path)
+    con.row_factory = sqlite3.Row
+    cur = con.cursor()
+
+    # Fetch all known finding IDs so we can warn about mismatches
+    cur.execute('SELECT total, findings_json FROM ivanti_findings_cache WHERE id = 1')
+    cache_row = cur.fetchone()
+    known_ids = set()
+    if cache_row and cache_row['findings_json']:
+        import json
+        try:
+            findings = json.loads(cache_row['findings_json'])
+            known_ids = {str(f['id']) for f in findings}
+        except Exception:
+            pass  # non-fatal — we'll still import and just skip the warning
+
+    # ----------------------------------------------------------------- process
+    inserted = 0
+    updated  = 0
+    skipped  = 0
+
+    for finding_id, note in rows:
+        str_id = str(finding_id)
+
+        if known_ids and str_id not in known_ids:
+            print(f'  WARNING: finding ID "{str_id}" not found in current cache — '
+                  f'note will be stored but won\'t display until a sync pulls that finding')
+
+        # Check if a note already exists
+        cur.execute('SELECT note FROM ivanti_finding_notes WHERE finding_id = ?', (str_id,))
+        existing = cur.fetchone()
+
+        if existing:
+            if existing['note'] == note:
+                print(f'  SKIP    {str_id} — note unchanged')
+                skipped += 1
+                continue
+            action = 'UPDATE'
+            updated += 1
+        else:
+            action = 'INSERT'
+            inserted += 1
+
+        print(f'  {action:6s}  {str_id}  →  {note[:80]}{"…" if len(note) > 80 else ""}')
+
+        if not args.dry_run:
+            cur.execute(
+                """
+                INSERT INTO ivanti_finding_notes (finding_id, note, updated_at)
+                VALUES (?, ?, datetime('now'))
+                ON CONFLICT(finding_id) DO UPDATE
+                SET note = excluded.note, updated_at = datetime('now')
+                """,
+                (str_id, note)
+            )
+
+    # ----------------------------------------------------------------- summary
+    print()
+    if args.dry_run:
+        print(f'DRY RUN complete — would insert {inserted}, update {updated}, skip {skipped}.')
+    else:
+        con.commit()
+        print(f'Done — inserted {inserted}, updated {updated}, skipped {skipped} (unchanged).')
+
+    con.close()
+
+
+if __name__ == '__main__':
+    run(parse_args())