diff --git a/backend/app/scoreboard_candidate_backfill.py b/backend/app/scoreboard_candidate_backfill.py
new file mode 100644
index 0000000..e604698
--- /dev/null
+++ b/backend/app/scoreboard_candidate_backfill.py
@@ -0,0 +1,176 @@
+"""Backfill public scoreboard candidates for RCON match link correlation.
+
+This command intentionally reuses the existing historical public-scoreboard
+archive ingestion path. The RCON materialized detail endpoint can only expose a
+safe public match URL when a matching scoreboard row exists in historical_matches.
+"""
+
+from __future__ import annotations
+
+import argparse
+import json
+from datetime import datetime, timezone
+from typing import Iterable
+
+from .historical_ingestion import run_incremental_refresh
+from .scoreboard_origins import list_trusted_public_scoreboard_origins
+
+DEFAULT_MAX_PAGES = 20
+DEFAULT_PAGE_SIZE = 100
+DEFAULT_DETAIL_WORKERS = 4
+
+
+def main(argv: Iterable[str] | None = None) -> int:
+    parser = build_arg_parser()
+    args = parser.parse_args(list(argv) if argv is not None else None)
+
+    from_timestamp = _parse_timestamp(args.from_timestamp, option_name="--from")
+    to_timestamp = _parse_timestamp(args.to_timestamp, option_name="--to")
+    if to_timestamp <= from_timestamp:
+        parser.error("--to must be later than --from")
+
+    _validate_server(args.server_slug, parser)
+    overlap_hours = _calculate_overlap_hours(from_timestamp)
+
+    progress_events: list[dict[str, object]] = []
+
+    def _capture_progress(payload: dict[str, object]) -> None:
+        progress_events.append(payload)
+        print(json.dumps(payload, ensure_ascii=False))
+
+    result = run_incremental_refresh(
+        server_slug=args.server_slug,
+        max_pages=args.max_pages,
+        page_size=args.page_size,
+        detail_workers=args.detail_workers,
+        overlap_hours=overlap_hours,
+        rebuild_snapshots=False,
+        progress_callback=_capture_progress,
+    )
+
+    totals = dict(result.get("totals") or {})
+    report = {
+        "status": result.get("status"),
+        "server": args.server_slug,
+        "requested_window": {
+            "from": _format_timestamp(from_timestamp),
+            "to": _format_timestamp(to_timestamp),
+        },
+        "ingestion_policy": {
+            "overlap_hours": overlap_hours,
+            "max_pages": args.max_pages,
+            "page_size": args.page_size,
+            "detail_workers": args.detail_workers,
+            "implementation": "historical-ingestion-public-scoreboard-candidates",
+        },
+        "candidates_seen": _coerce_int(totals.get("matches_seen")),
+        "candidates_inserted": _coerce_int(totals.get("matches_inserted")),
+        "candidates_updated": _coerce_int(totals.get("matches_updated")),
+        "player_rows_inserted": _coerce_int(totals.get("player_rows_inserted")),
+        "player_rows_updated": _coerce_int(totals.get("player_rows_updated")),
+        "skipped_unsafe_urls": 0,
+        "errors": _extract_errors(progress_events, result),
+        "raw_result": result,
+    }
+    print(json.dumps(report, ensure_ascii=False, indent=2))
+    return 0 if str(result.get("status")) == "ok" else 1
+
+
+def build_arg_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(
+        description="Backfill public scoreboard match candidates for RCON link correlation.",
+    )
+    parser.add_argument(
+        "--server",
+        dest="server_slug",
+        required=True,
+        help="trusted historical server slug, e.g. comunidad-hispana-02",
+    )
+    parser.add_argument(
+        "--from",
+        dest="from_timestamp",
+        required=True,
+        help="inclusive UTC-ish ISO timestamp used as the lower backfill bound",
+    )
+    parser.add_argument(
+        "--to",
+        dest="to_timestamp",
+        required=True,
+        help="exclusive UTC-ish ISO timestamp kept in the report for traceability",
+    )
+    parser.add_argument(
+        "--max-pages",
+        type=int,
+        default=DEFAULT_MAX_PAGES,
+        help=f"maximum public-scoreboard pages to scan, default {DEFAULT_MAX_PAGES}",
+    )
+    parser.add_argument(
+        "--page-size",
+        type=int,
+        default=DEFAULT_PAGE_SIZE,
+        help=f"public-scoreboard page size, default {DEFAULT_PAGE_SIZE}",
+    )
+    parser.add_argument(
+        "--detail-workers",
+        type=int,
+        default=DEFAULT_DETAIL_WORKERS,
+        help=f"parallel detail workers, default {DEFAULT_DETAIL_WORKERS}",
+    )
+    return parser
+
+
+def _validate_server(server_slug: str, parser: argparse.ArgumentParser) -> None:
+    trusted_slugs = {origin.slug for origin in list_trusted_public_scoreboard_origins()}
+    if server_slug not in trusted_slugs:
+        parser.error(
+            f"unknown or untrusted server '{server_slug}'. "
+            f"Allowed values: {', '.join(sorted(trusted_slugs))}"
+        )
+
+
+def _parse_timestamp(value: str, *, option_name: str) -> datetime:
+    try:
+        parsed = datetime.fromisoformat(value.strip().replace("Z", "+00:00"))
+    except ValueError as exc:
+        raise argparse.ArgumentTypeError(
+            f"{option_name} must be an ISO timestamp, got {value!r}"
+        ) from exc
+    if parsed.tzinfo is None:
+        parsed = parsed.replace(tzinfo=timezone.utc)
+    return parsed.astimezone(timezone.utc)
+
+
+def _calculate_overlap_hours(from_timestamp: datetime) -> int:
+    now = datetime.now(timezone.utc)
+    delta_seconds = max(0, int((now - from_timestamp).total_seconds()))
+    # Add one hour so the requested lower bound is safely included after integer rounding.
+    return max(1, (delta_seconds // 3600) + 1)
+
+
+def _format_timestamp(value: datetime) -> str:
+    return value.astimezone(timezone.utc).isoformat().replace("+00:00", "Z")
+
+
+def _coerce_int(value: object) -> int:
+    try:
+        return int(value or 0)
+    except (TypeError, ValueError):
+        return 0
+
+
+def _extract_errors(
+    progress_events: list[dict[str, object]],
+    result: dict[str, object],
+) -> list[dict[str, object]]:
+    errors: list[dict[str, object]] = []
+    if str(result.get("status")) != "ok":
+        errors.append({"scope": "result", "message": str(result)})
+    for event in progress_events:
+        text = str(event.get("event") or "")
+        if "error" in text or "failed" in text:
+            errors.append(event)
+    return errors
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())