diff --git a/backend/services/fetchers/acars_summarize.py b/backend/services/fetchers/acars_summarize.py
new file mode 100644
index 0000000..8d45b0c
--- /dev/null
+++ b/backend/services/fetchers/acars_summarize.py
@@ -0,0 +1,570 @@
+"""Heuristics to summarize ACARS/VDL payloads across airlines for dossier display."""
+
+from __future__ import annotations
+
+import re
+from typing import Any
+
+# --- shared patterns ---
+
+_ICAO_AIRPORT = re.compile(r"\b([A-Z]{4})\b")
+_TAIL = re.compile(r"\b(?:N[0-9A-Z]{3,6}|G-[A-Z]{4,5}|[A-Z]-[A-Z]{4,5}|[A-Z]{2}-[A-Z]{3,4})\b")
+# Major carriers — explicit list avoids matching FL280, GS450, etc.
+_FLIGHT = re.compile(
+    r"\b(?:"
+    r"WN|SWA|UA|UAL|AA|AAL|DL|DAL|AS|ASA|B6|JBU|NK|NKS|F9|FFT|G4|HA|HAL|SY|MX|"
+    r"FDX|UPS|GTI|ABX|ATN|RCH|CNV|EVAC|SAM|REACH|"
+    r"BA|BAW|AF|AFR|LH|DLH|KL|KLM|QF|QFA|EK|UAE|QR|QTR|TK|THY|AC|ACA|WS|WJA|"
+    r"FR|RYR|U2|EZY|VS|VIR|NH|ANA|JL|JAL|CX|CPA|SQ|SIA|NZ|ANZ|"
+    r"UA|CO|NW|US|HP|TW|VX|AS|QX|OO|YX|MQ|OH|9E|"
+    r"JT|JSA|VA|VOZ|NZ|QF|EK|ET|MS|SU|LO|SK|AY|IB|UX|TP|TAP"
+    r")\d{1,5}\b",
+    re.I,
+)
+# IATA flight numbers on FI lines and standalone (e.g. UO614, CX889).
+_FI_FLIGHT = re.compile(r"\b([A-Z]{2,3}\d{1,4})\b")
+_NON_FLIGHT_TOKENS = frozenset(
+    {"FL", "FT", "GS", "KT", "RW", "NM", "TD", "TO", "ON", "IN", "OF", "AT", "DA", "AA", "AD"}
+)
+_FI_BLOCK = re.compile(
+    r"FI\s+([A-Z0-9]{2,5}\d{1,5})"
+    r"(?:/AN\s+([A-Z0-9\-]+))?"
+    r"(?:/DA\s+([A-Z]{4}))?"
+    r"(?:/(?:AA|AD|DS)\s+([A-Z]{4}))?",
+    re.I,
+)
+_AC_TYPE = re.compile(
+    r"\b(?:B\d{3,4}(?:-\d{3}|MAX|ER|LR|F)?|A\d{3,4}(?:-\d{3}|NEO|LR)?|"
+    r"E\d{3}|MD-\d{2}|DC-\d{2}|B77[0-9LWR]?|B78[79]|A35[09]|A33[0-9]|CRJ\d{2,3}|E\d{3})\b",
+    re.I,
+)
+
+# --- message family patterns ---
+
+_TRACK_HEADER = re.compile(
+    r"^\+\+86501,([^,]+),([^,]+),(\d{6}),([^,]+),([A-Z]{4}),([A-Z]{4})",
+    re.I,
+)
+_POS_HEADER = re.compile(r"^POS(N?\d{4,5}[NS]?\d{4,5}[EW]?)", re.I)
+_POS_COORDS = re.compile(r"^N?(\d{4,5})([NS])(\d{4,6})([EW])", re.I)
+_WAYPOINT = re.compile(
+    r"^(?:N)?(\d{1,2}\d{2}\.\d),W(\d{1,3}\d{2}\.\d),(\d{6}),(\d+),",
+    re.I,
+)
+_PERF_HEADER = re.compile(
+    r"^[\w]+,(\d+),([^,]+),(\d{6}),([^,]+),([A-Z]{4}),([A-Z]{4})",
+    re.I,
+)
+_PHASE_SNAPSHOT = re.compile(
+    r"^(\d{2}\.\d{2}\.\d{2}),(CL|CR|DE|TO|LD|ER|GND),[^,]*,[^,]*,[^,]*,[^,]*,[^,]*,[^,]*,"
+    r"N(\d+)\.(\d+),W(\d+)\.(\d+)",
+    re.I,
+)
+_TRAJECTORY_HEADER = re.compile(r"^76401\s*$", re.I)
+_TRAJECTORY_ROUTE = re.compile(r"^02E24([A-Z]{4})([A-Z]{4})\s*$", re.I)
+_COMPRESSED_WP = re.compile(r"^N(\d{5})W(\d{5})", re.I)
+_FPN = re.compile(r"^FPN/?", re.I)
+_OOOI_TIMES = re.compile(r"\b(OUT|OFF|ON|IN)\s*(\d{4,6})\b", re.I)
+_OOOI_STATUS = re.compile(r"\b(OUT|OFF|ON|IN)\s*,\s*(LO|CL|ON|OF|CLOS)\b", re.I)
+_ETA = re.compile(r"\bETA\s+(\d{3,4}Z?)\b", re.I)
+_DEP_ARR = re.compile(r"^(DEP|ARR|DLA|ALR)\b", re.I)
+_WX = re.compile(r"^(?:WXR?\d*|WX\s|MET\b|/WX\b)", re.I)
+_REQ = re.compile(r"^(?:REQ|REQUEST)\b", re.I)
+_LDR = re.compile(r"^LDR\d+", re.I)
+_PIREP = re.compile(r"^#(?:CFB|DFB)", re.I)
+_ATN = re.compile(r"^USADCXA\.AT1\.", re.I)
+_CPDLC = re.compile(r"^(?:DM-|UM-|AT1\.|ATC\s)", re.I)
+_ENG = re.compile(r"^(?:ENG\d|/ENG|OILTEMP|EGT\b)", re.I)
+_DOOR = re.compile(r"^(?:DOOR|CABIN|SMOKE)\b", re.I)
+_VDL_FRAME = re.compile(r"^[0-9A-F]{6,8}[A-Z]?\s*$", re.I)
+_FRAGMENT = re.compile(r"^[,0\s]+(?:,\d{5,8},\d{5,8},\d{5,8})*$", re.I)
+_GARBLED_VDL = re.compile(r"[)Z][A-Z0-9,\-:]{20,}")
+_MOSTLY_OPAQUE = re.compile(r"^[0-9A-Fa-f\s.\-+/,]{40,}$")
+_FREE_TEXT_POS = re.compile(
+    r"^POS\s+N?(\d{1,2}\.\d+)\s+([NS])\s+W?(\d{1,3}\.\d+)\s+([EW])\s+FL(\d{3})",
+    re.I,
+)
+_CLIMB_REQ = re.compile(r"\b(?:CLIMB|DESCEND|REQUEST)\s+(?:FL)?(\d{2,3})\b", re.I)
+
+_LABEL_HINTS: dict[str, str] = {
+    "00": "out (gate)",
+    "01": "off (takeoff)",
+    "02": "on (landing)",
+    "03": "in (gate)",
+    "10": "position",
+    "15": "waypoint",
+    "20": "position",
+    "40": "ops / clearance",
+    "44": "OOOI + position",
+    "80": "weather",
+    "81": "wind",
+    "B1": "engine 1",
+    "B2": "engine 2",
+    "B3": "engine 3",
+    "B4": "engine 4",
+    "M1": "maintenance",
+    "M2": "maintenance",
+    "M3": "maintenance",
+    "M4": "maintenance",
+    "Q0": "position / OOOI",
+    "H1": "terminal",
+    "D0": "ATC clearance",
+    "S1": "system status",
+    "SA": "system status",
+    "SB": "system status",
+    "4T": "met report",
+    "5Z": "free text",
+}
+
+
+def _result(
+    summary: str,
+    *,
+    kind: str,
+    readable: bool = True,
+    hidden: bool = False,
+) -> dict[str, Any]:
+    return {
+        "summary": summary,
+        "kind": kind,
+        "readable": readable,
+        "hidden": hidden,
+    }
+
+
+def _phase_name(code: str) -> str:
+    return {
+        "CL": "climb",
+        "CR": "cruise",
+        "DE": "descent",
+        "ER": "en route",
+        "TO": "takeoff",
+        "LD": "landed",
+        "ON": "on ground",
+        "OF": "off block",
+        "GND": "on ground",
+        "LO": "level",
+    }.get(code.upper(), code.upper() or "unknown")
+
+
+def _fmt_coords(lat_deg: str, lat_frac: str, lon_deg: str, lon_frac: str) -> str:
+    return f"{int(lat_deg)}°{lat_frac}'N {int(lon_deg)}°{lon_frac}'W"
+
+
+def _parse_pos_coords(token: str) -> str | None:
+    token = token.upper().lstrip("POS")
+    match = _POS_COORDS.match(token)
+    if not match:
+        return None
+    lat, lat_dir, lon, lon_dir = match.groups()
+    lat_v = f"{int(lat[:2])}°{lat[2:]}.{lat[4:] if len(lat) > 4 else '0'}'{lat_dir}"
+    lon_v = f"{int(lon[:3])}°{lon[3:]}.{lon[5:] if len(lon) > 5 else '0'}'{lon_dir}"
+    return f"{lat_v} {lon_v}"
+
+
+def _extract_route(raw: str) -> str:
+    fi = _FI_BLOCK.search(raw)
+    if fi:
+        flight, _tail, dep, dest = fi.groups()
+        parts = [flight.upper()]
+        if dep and dest:
+            parts.append(f"{dep}→{dest}")
+        elif dep:
+            parts.append(f"from {dep}")
+        elif dest:
+            parts.append(f"to {dest}")
+        return " · ".join(parts)
+
+    airports = _ICAO_AIRPORT.findall(raw)
+    # Filter duplicates while preserving order
+    seen: set[str] = set()
+    ordered: list[str] = []
+    for apt in airports:
+        if apt in seen:
+            continue
+        seen.add(apt)
+        ordered.append(apt)
+    if len(ordered) >= 2:
+        return f"{ordered[0]}→{ordered[-1]}"
+    if ordered:
+        return ordered[0]
+    return ""
+
+
+def _extract_flight(raw: str) -> str:
+    fi = _FI_BLOCK.search(raw)
+    if fi and fi.group(1):
+        return fi.group(1).upper()
+    for match in _FLIGHT.finditer(raw):
+        return match.group(0).upper()
+    for match in _FI_FLIGHT.finditer(raw):
+        token = match.group(1).upper()
+        prefix = re.match(r"^([A-Z]+)", token)
+        if prefix and prefix.group(1) not in _NON_FLIGHT_TOKENS:
+            return token
+    return ""
+
+
+def _extract_phase_snapshot(raw: str) -> str | None:
+    for line in raw.splitlines():
+        match = _PHASE_SNAPSHOT.match(line.strip())
+        if not match:
+            continue
+        time_s, phase, lat_d, lat_f, lon_d, lon_f = match.groups()
+        coords = _fmt_coords(lat_d, lat_f, lon_d, lon_f)
+        return f"{_phase_name(phase)} · {coords} · {time_s}Z"
+    return None
+
+
+def _has_aircraft_context(raw: str) -> bool:
+    head = raw[:160].upper()
+    if _AC_TYPE.search(head):
+        return True
+    if _FLIGHT.search(head):
+        return True
+    if _FI_BLOCK.search(head):
+        return True
+    return False
+
+
+def _is_fragment(raw: str) -> bool:
+    first = raw.splitlines()[0].strip()
+    if _FRAGMENT.match(first):
+        return True
+    if re.match(r"^[,0]{1,12}$", first):
+        return True
+    if first.startswith("000000") or first.startswith(",000000"):
+        return True
+    if re.match(r"^\d{2}\.\d{2}\.\d{2},", first) and not _has_aircraft_context(raw):
+        return True
+    return False
+
+
+def _summarize_oooi(raw: str, label: str) -> dict[str, Any] | None:
+    times = _OOOI_TIMES.findall(raw)
+    statuses = _OOOI_STATUS.findall(raw)
+    if not times and not statuses and label not in {"00", "01", "02", "03", "44", "Q0"}:
+        return None
+
+    events: list[str] = []
+    for event, value in times:
+        events.append(f"{event.upper()} {value}")
+    for event, status in statuses:
+        events.append(f"{event.upper()} ({_phase_name(status)})")
+
+    if label in {"00", "01", "02", "03"} and not events:
+        events.append(_LABEL_HINTS[label])
+
+    if not events and "ON  ,LO" not in raw and "OFF,OFF" not in raw:
+        return None
+
+    route = _extract_route(raw)
+    flight = _extract_flight(raw)
+    prefix = "OOOI"
+    if label in _LABEL_HINTS:
+        prefix = f"OOOI ({_LABEL_HINTS[label]})"
+    bits = [prefix]
+    if flight:
+        bits.append(flight)
+    if route:
+        bits.append(route)
+    if events:
+        bits.append(", ".join(events[:4]))
+    return _result(" · ".join(bits), kind="oooi")
+
+
+def _summarize_position(raw: str, first_line: str) -> dict[str, Any] | None:
+    upper = first_line.upper()
+    pos_token = re.search(r"POSN?\d", raw, re.I)
+    if not (upper.startswith("POS") or _POS_HEADER.match(first_line) or pos_token):
+        return None
+
+    coord_line = first_line
+    if pos_token and not upper.startswith("POS"):
+        coord_line = raw[pos_token.start() :].split(",")[0]
+
+    coords = _parse_pos_coords(coord_line)
+    free = _FREE_TEXT_POS.match(raw)
+    fl = ""
+    if free:
+        lat, lat_dir, lon, lon_dir, fl = free.groups()
+        coords = f"{lat}°{lat_dir} {lon}°{lon_dir}"
+        fl = f"FL{fl}"
+
+    route = _extract_route(raw)
+    flight = _extract_flight(raw)
+    parts = ["Position report"]
+    if flight:
+        parts.append(flight)
+    if route:
+        parts.append(route)
+    if coords:
+        parts.append(coords)
+    if fl:
+        parts.append(fl)
+    elif re.search(r"\bFL?\d{3}\b", raw):
+        fl_match = re.search(r"\bFL?(\d{2,3})\b", raw)
+        if fl_match:
+            parts.append(f"FL{fl_match.group(1)}")
+    return _result(" · ".join(parts), kind="position")
+
+
+def _summarize_performance(raw: str, first_line: str) -> dict[str, Any] | None:
+    match = _PERF_HEADER.match(first_line)
+    if not match or not _AC_TYPE.search(first_line):
+        return None
+
+    _serial, ac_type, _date, flight, dep, dest = match.groups()
+    phase_bits = _extract_phase_snapshot(raw) or ""
+    extra = f" · {phase_bits}" if phase_bits else ""
+
+    if "FHP" in raw or "SIN," in raw or "SOU," in raw:
+        title, kind = "Engine health (FHP)", "engine_health"
+    elif "OATTO" in raw or "LPACKCL" in raw or "RPACKCL" in raw:
+        title, kind = "Pack temperature", "pack_temp"
+    elif "FLAPS" in raw.upper():
+        title, kind = "Climb performance", "climb_perf"
+    elif "FRE," in raw or "FEX," in raw:
+        title, kind = "Fuel/performance snapshot", "fuel_perf"
+    else:
+        title, kind = "Flight performance", "performance"
+
+    return _result(
+        f"{title} · {flight} · {ac_type} · {dep}→{dest}{extra}",
+        kind=kind,
+    )
+
+
+def _summarize_by_label(label: str, raw: str, first_line: str) -> dict[str, Any] | None:
+    label_u = label.upper()
+    hint = _LABEL_HINTS.get(label_u, "")
+
+    if label_u in {"B1", "B2", "B3", "B4"} or _ENG.match(first_line):
+        eng = label_u if label_u.startswith("B") else "Engine"
+        return _result(f"{eng} data report", kind="engine", readable=bool(hint))
+
+    if label_u.startswith("M") and label_u[1:2].isdigit():
+        return _result(f"Maintenance ({hint or 'system report'})", kind="maintenance")
+
+    if label_u in {"80", "81", "4T"} or _WX.match(first_line):
+        apt = _ICAO_AIRPORT.search(raw)
+        apt_s = f" · {apt.group(1)}" if apt else ""
+        return _result(f"Weather report{apt_s}", kind="weather")
+
+    if label_u == "D0" or _REQ.match(first_line) or _CLIMB_REQ.search(raw):
+        climb = _CLIMB_REQ.search(raw)
+        if climb:
+            return _result(f"Altitude request · FL{climb.group(1)}", kind="request")
+        return _result("ATC / ops request", kind="request")
+
+    if label_u in {"40", "5Z"} and len(raw) < 200:
+        text = raw.replace("\n", " · ")[:140]
+        return _result(f"Ops message · {text}", kind="ops")
+
+    return None
+
+
+def summarize_datalink_message(
+    *,
+    label: str = "",
+    text: str = "",
+    source_type: str = "",
+) -> dict[str, Any]:
+    """Return {summary, kind, readable, hidden} for a cached datalink message."""
+    raw = (text or "").strip()
+    if not raw:
+        return _result("", kind="empty", readable=False, hidden=True)
+
+    first_line = raw.splitlines()[0].strip()
+    upper = first_line.upper()
+    label_u = label.upper()
+
+    if _is_fragment(raw):
+        return _result(
+            "Split telemetry fragment (part of a longer VDL message)",
+            kind="fragment",
+            readable=False,
+            hidden=True,
+        )
+
+    if _ATN.match(first_line) or _CPDLC.match(first_line):
+        tail = _TAIL.search(raw)
+        return _result(
+            "Datalink protocol / CPDLC header" + (f" · {tail.group(0)}" if tail else ""),
+            kind="protocol",
+            readable=False,
+            hidden=True,
+        )
+
+    if label_u == "37" or (_VDL_FRAME.match(first_line) and len(raw) < 160):
+        if _GARBLED_VDL.search(raw) or len(raw) < 160:
+            return _result("VDL binary frame (undecoded)", kind="vdl_binary", readable=False, hidden=True)
+
+    # --- structured families (order matters) ---
+
+    if _DEP_ARR.match(first_line):
+        kind_word = first_line.split()[0].upper()
+        route = _extract_route(raw)
+        flight = _extract_flight(raw)
+        title = {"DEP": "Departure", "ARR": "Arrival", "DLA": "Delay", "ALR": "Alert"}.get(
+            kind_word, kind_word
+        )
+        bits = [title]
+        if flight:
+            bits.append(flight)
+        if route:
+            bits.append(route)
+        return _result(" · ".join(bits), kind=kind_word.lower())
+
+    oooi = _summarize_oooi(raw, label_u)
+    if oooi:
+        return oooi
+
+    match = _TRACK_HEADER.match(first_line)
+    if match:
+        tail, ac_type, _date, flight, dep, dest = match.groups()
+        lines = [line.strip() for line in raw.splitlines() if line.strip()]
+        waypoint_lines = [line for line in lines if _WAYPOINT.match(line.lstrip("N"))]
+        phase = ""
+        if waypoint_lines:
+            parts = waypoint_lines[-1].rstrip(",").split(",")
+            if len(parts) >= 8:
+                phase = _phase_name(parts[7])
+        wp_count = len(waypoint_lines) or max(0, len(lines) - 2)
+        summary = (
+            f"Track report · {flight} · {tail} ({ac_type}) · {dep}→{dest}"
+            + (f" · {wp_count} waypoint(s)" + (f" · {phase}" if phase else ""))
+        )
+        return _result(summary, kind="track")
+
+    pos = _summarize_position(raw, first_line)
+    if pos:
+        return pos
+
+    if _FPN.match(first_line):
+        route = _extract_route(raw)
+        flight = _extract_flight(raw)
+        bits = ["Flight plan"]
+        if flight:
+            bits.append(flight)
+        if route:
+            bits.append(route)
+        return _result(" · ".join(bits), kind="flight_plan")
+
+    if _PIREP.match(first_line):
+        return _result("Pilot report (PIREP)", kind="pirep")
+
+    lines = [line.strip() for line in raw.splitlines() if line.strip()]
+    if _TRAJECTORY_HEADER.match(first_line) or (len(lines) >= 2 and _TRAJECTORY_ROUTE.match(lines[1])):
+        route = ""
+        route_match = next((m for line in lines if (m := _TRAJECTORY_ROUTE.match(line))), None)
+        if route_match:
+            route = f" · {route_match.group(1)}→{route_match.group(2)}"
+        wp_count = sum(1 for line in lines if _COMPRESSED_WP.match(line))
+        return _result(
+            f"Trajectory / ADS report{route}" + (f" · {wp_count} point(s)" if wp_count else ""),
+            kind="trajectory",
+        )
+
+    perf = _summarize_performance(raw, first_line)
+    if perf:
+        return perf
+
+    if _LDR.match(first_line):
+        route = _extract_route(raw)
+        return _result(f"Load report · {route}" if route else "Load report", kind="load")
+
+    if _WX.match(first_line):
+        route = _extract_route(raw)
+        return _result(f"Weather request · {route or 'en route'}", kind="weather")
+
+    if _DOOR.match(first_line):
+        return _result("Cabin / door advisory", kind="cabin")
+
+    if _WAYPOINT.match(first_line.lstrip("N")):
+        parts = first_line.lstrip("N").rstrip(",").split(",")
+        if len(parts) >= 4:
+            lat, lon, _t, alt = parts[0], parts[1], parts[2], parts[3]
+            phase = _phase_name(parts[7]) if len(parts) >= 8 else ""
+            summary = f"Waypoint · {lat},{lon} · alt {alt} ft" + (f" · {phase}" if phase else "")
+            return _result(summary, kind="waypoint")
+
+    label_summary = _summarize_by_label(label_u, raw, first_line)
+    if label_summary:
+        return label_summary
+
+    flight = _extract_flight(raw)
+    route = _extract_route(raw)
+    if flight and route:
+        return _result(f"Datalink · {flight} · {route}", kind="flight")
+
+    eta = _ETA.search(raw)
+    if eta and flight:
+        return _result(f"ETA update · {flight} · {eta.group(1)}", kind="eta")
+
+    if len(raw) < 100 and not _MOSTLY_OPAQUE.match(raw) and not _GARBLED_VDL.search(raw):
+        clean = raw.replace("\n", " · ")
+        if label_u in _LABEL_HINTS:
+            return _result(f"{_LABEL_HINTS[label_u].title()} · {clean}", kind="short")
+        return _result(clean, kind="short")
+
+    digit_ratio = sum(ch.isdigit() for ch in raw) / max(len(raw), 1)
+    if digit_ratio > 0.55 or _MOSTLY_OPAQUE.match(raw.replace(" ", "")) or _GARBLED_VDL.search(raw):
+        return _result(
+            "Binary / proprietary telemetry (undecoded)",
+            kind="vdl_binary",
+            readable=False,
+            hidden=True,
+        )
+
+    if label_u in _LABEL_HINTS:
+        return _result(
+            f"{_LABEL_HINTS[label_u].title()} message",
+            kind=label_u.lower(),
+            readable=False,
+            hidden=False,
+        )
+
+    return _result(
+        first_line[:100] + ("…" if len(first_line) > 100 else ""),
+        kind="raw",
+        readable=False,
+        hidden=False,
+    )
+
+
+def prepare_datalink_display(messages: list[dict[str, Any]]) -> dict[str, Any]:
+    """Attach summaries and filter noise for dossier display."""
+    enriched: list[dict[str, Any]] = []
+    hidden_count = 0
+    seen_summaries: set[str] = set()
+
+    for message in messages:
+        meta = summarize_datalink_message(
+            label=str(message.get("label") or ""),
+            text=str(message.get("text") or ""),
+            source_type=str(message.get("source_type") or ""),
+        )
+        item = {**message, **meta}
+        if item.get("hidden"):
+            hidden_count += 1
+            continue
+
+        # Drop back-to-back duplicate summaries (common with multi-part VDL)
+        sig = f"{item.get('kind')}|{item.get('summary')}"
+        if sig in seen_summaries and item.get("kind") not in {"short", "ops", "request"}:
+            hidden_count += 1
+            continue
+        seen_summaries.add(sig)
+
+        enriched.append(item)
+
+    return {
+        "messages": enriched,
+        "hidden_count": hidden_count,
+        "total_count": len(messages),
+    }
+
+
+def attach_summaries(messages: list[dict[str, Any]]) -> list[dict[str, Any]]:
+    return prepare_datalink_display(messages)["messages"]
diff --git a/backend/services/fetchers/airframes.py b/backend/services/fetchers/airframes.py
index 9d1f3cc..3d47b41 100644
--- a/backend/services/fetchers/airframes.py
+++ b/backend/services/fetchers/airframes.py
@@ -593,9 +593,15 @@ def lookup_datalink_messages(
         if queued_refresh:
             _ensure_stagger_worker()
 
+    from services.fetchers.acars_summarize import prepare_datalink_display
+
+    display = prepare_datalink_display(messages)
+
     return {
         "configured": True,
-        "messages": messages,
+        "messages": display["messages"],
+        "hidden_count": display["hidden_count"],
+        "total_count": display["total_count"],
         "last_success_at": last_success_at,
         "queued_refresh": queued_refresh,
         "priority_scan": queued_refresh,
diff --git a/backend/tests/test_acars_summarize.py b/backend/tests/test_acars_summarize.py
new file mode 100644
index 0000000..78ed5a4
--- /dev/null
+++ b/backend/tests/test_acars_summarize.py
@@ -0,0 +1,182 @@
+from services.fetchers.acars_summarize import prepare_datalink_display, summarize_datalink_message
+
+# --- Southwest (existing) ---
+
+
+def test_summarize_track_report():
+    text = """++86501,N8997Q,B7378MAX,260620,WN3743,KMSP,KMDW,0496,SMX34-2502-F320
+6
+N4432.0,W09305.6,201041,15193,-08.3,310,044,CL,00000,0,"""
+    meta = summarize_datalink_message(label="H1", text=text, source_type="vdl")
+    assert meta["kind"] == "track"
+    assert "WN3743" in meta["summary"]
+    assert "KMSP→KMDW" in meta["summary"]
+
+
+def test_summarize_sw_performance_cruise():
+    text = """72740,7852,B737-700,260624,WN0120,KABQ,KDEN,1986,SW2501
+18.45.14,CR,1575,28981,280.0,.729,-32.3,-06.5,N3601.3,W10655.7,131240
+   0.48,FHP,AIR
+SIN,-1.42  0.30  0.29"""
+    meta = summarize_datalink_message(label="H1", text=text, source_type="vdl")
+    assert meta["kind"] == "engine_health"
+    assert "WN0120" in meta["summary"]
+    assert "cruise" in meta["summary"]
+
+
+def test_summarize_sw_climb_performance():
+    text = """05201,7852,B737-700,260624,WN0120,KABQ,KDEN,1986,SW2501
+18.38.08,CL,1149,15631,257.0,.520,000.0,014.5,N3515.4,W10649.4,132800
+001.40,001,4100,FLAPS-UP"""
+    meta = summarize_datalink_message(label="H1", text=text, source_type="vdl")
+    assert meta["kind"] == "climb_perf"
+    assert "climb" in meta["summary"]
+
+
+def test_summarize_trajectory():
+    text = """76401
+02E24KABQKDEN
+N35112W10679318361096P014343008G000022::I0:9W
+N35195W10681118371370P006337009G000022::Q0OXW"""
+    meta = summarize_datalink_message(label="H1", text=text, source_type="vdl")
+    assert meta["kind"] == "trajectory"
+    assert "KABQ→KDEN" in meta["summary"]
+
+
+def test_fragment_hidden():
+    text = "0000000,00000000,00000000\n18.38.23,16395,250.2,.510,01.07,01.04,00,00000000"
+    meta = summarize_datalink_message(label="H1", text=text, source_type="vdl")
+    assert meta["kind"] == "fragment"
+    assert meta["hidden"] is True
+
+
+def test_vdl_binary_hidden():
+    text = "014F63N\n)AJQZ)LC0Z0IP-M7O,ZHN3-M,73ZO,UU-ZOS1Z7PPZMSN1ZN"
+    meta = summarize_datalink_message(label="37", text=text, source_type="vdl")
+    assert meta["hidden"] is True
+
+
+# --- United / Delta / American ---
+
+
+def test_united_free_text_position():
+    text = "POS N40.123 W074.456 FL350 GS450 1425Z"
+    meta = summarize_datalink_message(label="Q0", text=text, source_type="vdl")
+    assert meta["kind"] == "position"
+    assert "FL350" in meta["summary"]
+
+
+def test_delta_oooi_out():
+    text = "OUT 1425 12JAN KATL"
+    meta = summarize_datalink_message(label="00", text=text, source_type="acars")
+    assert meta["kind"] == "oooi"
+    assert "OUT 1425" in meta["summary"]
+
+
+def test_american_fi_block():
+    text = "FI AA100/AN N100AA/DA KDFW/AA KLAX OUT 1832 OFF 1845"
+    meta = summarize_datalink_message(label="44", text=text, source_type="acars")
+    assert "AA100" in meta["summary"]
+    assert "KDFW→KLAX" in meta["summary"]
+
+
+def test_united_performance_a320():
+    text = """88401,4521,A320-200,260624,UA1234,KORD,KDEN,1200,UA2501
+19.10.22,CR,2200,35000,450.0,.820,-45.0,-02.0,N3950.1,W10440.2,125000"""
+    meta = summarize_datalink_message(label="H1", text=text, source_type="vdl")
+    assert meta["kind"] == "performance"
+    assert "UA1234" in meta["summary"]
+    assert "KORD→KDEN" in meta["summary"]
+
+
+# --- International ---
+
+
+def test_british_airways_engine():
+    text = "ENG1 N1 92.5 N2 95.1 EGT 512 FF 2850"
+    meta = summarize_datalink_message(label="B1", text=text, source_type="satcom")
+    assert meta["kind"] == "engine"
+
+
+def test_qantas_fi_position():
+    text = "FI QF9/AN VH-OQA/DA YSSY/AD EGLL POSN32249E045047,,082806,380,DEBNI"
+    meta = summarize_datalink_message(label="H1", text=text, source_type="acars")
+    assert meta["kind"] == "position"
+    assert "QF9" in meta["summary"]
+    assert "YSSY→EGLL" in meta["summary"]
+
+
+def test_lufthansa_weather():
+    text = "WX 250/045 SAT -42 TB MOD EDDF"
+    meta = summarize_datalink_message(label="80", text=text, source_type="acars")
+    assert meta["kind"] == "weather"
+    assert "EDDF" in meta["summary"]
+
+
+def test_air_france_request():
+    text = "REQUEST FL370 DUE TURB"
+    meta = summarize_datalink_message(label="H1", text=text, source_type="vdl")
+    assert meta["kind"] == "request"
+    assert "FL370" in meta["summary"]
+
+
+# --- Cargo / military ---
+
+
+def test_fedex_flight():
+    text = "++86501,N123FE,B763,260624,FDX1544,KMEM,KORD,0498,SMX34"
+    meta = summarize_datalink_message(label="H1", text=text, source_type="vdl")
+    assert meta["kind"] == "track"
+    assert "FDX1544" in meta["summary"]
+
+
+def test_military_rch():
+    text = "POSN3840.5 W07720.1 FL280 RCH123 KADW KDMA"
+    meta = summarize_datalink_message(label="Q0", text=text, source_type="acars")
+    assert "RCH123" in meta["summary"]
+
+
+# --- Ops / misc ---
+
+
+def test_flight_plan():
+    text = "FPN/RI:DA:KJFK:AA:EGLL..MERIT:D:MERIT"
+    meta = summarize_datalink_message(label="H1", text=text, source_type="vdl")
+    assert meta["kind"] == "flight_plan"
+    assert "KJFK→EGLL" in meta["summary"]
+
+
+def test_departure_report():
+    text = "DEP FI DL456/DA KATL/AD KLAX OUT 1205"
+    meta = summarize_datalink_message(label="40", text=text, source_type="acars")
+    assert meta["kind"] == "dep"
+    assert "DL456" in meta["summary"]
+
+
+def test_pirep():
+    text = "#CFB/PIREP MOD TURB FL280 N3845 W09030"
+    meta = summarize_datalink_message(label="H1", text=text, source_type="acars")
+    assert meta["kind"] == "pirep"
+
+
+def test_prepare_filters_hidden_and_dedupes():
+    messages = [
+        {"id": 1, "label": "H1", "text": "POSN35259W106517,KABQ,KDEN", "source_type": "vdl"},
+        {"id": 2, "label": "H1", "text": "0000000,00000000,00000000", "source_type": "vdl"},
+        {"id": 3, "label": "37", "text": "014F63N\n)AJQZ)LC0Z", "source_type": "vdl"},
+        {
+            "id": 4,
+            "label": "H1",
+            "text": "72740,7852,B737-700,260624,WN0120,KABQ,KDEN\n18.45.14,CR,1575,28981",
+            "source_type": "vdl",
+        },
+        {
+            "id": 5,
+            "label": "H1",
+            "text": "72740,7852,B737-700,260624,WN0120,KABQ,KDEN\n18.45.14,CR,1575,28981",
+            "source_type": "vdl",
+        },
+    ]
+    display = prepare_datalink_display(messages)
+    assert display["hidden_count"] == 3
+    assert len(display["messages"]) == 2
diff --git a/frontend/src/components/DatalinkMessagesBlock.tsx b/frontend/src/components/DatalinkMessagesBlock.tsx
index 07774d1..7bede4d 100644
--- a/frontend/src/components/DatalinkMessagesBlock.tsx
+++ b/frontend/src/components/DatalinkMessagesBlock.tsx
@@ -9,11 +9,49 @@ type DatalinkMessage = {
   label?: string;
   text?: string;
   source_type?: string;
+  summary?: string;
+  kind?: string;
+  readable?: boolean;
 };
 
 const PRIORITY_POLL_MS = 3_000;
 const PRIORITY_POLL_MAX_MS = 45_000;
 
+function DatalinkMessageRow({ message }: { message: DatalinkMessage }) {
+  const [showRaw, setShowRaw] = useState(false);
+  const summary = message.summary?.trim();
+  const raw = message.text?.trim() || '';
+  const hasSummary = Boolean(summary);
+  const showRawBlock = showRaw || (!hasSummary && raw);
+
+  return (
+    <div className="text-[10px] font-mono leading-snug border border-[var(--border-primary)]/60 bg-black/20 px-2 py-1.5">
+      <div className="flex items-center gap-2 text-[var(--text-muted)] mb-0.5">
+        <span>{formatDatalinkTime(message.timestamp)}</span>
+        {message.label ? <span className="text-orange-400/90">{message.label}</span> : null}
+        {message.source_type ? <span className="truncate">{message.source_type}</span> : null}
+      </div>
+      {hasSummary ? (
+        <div className="text-[var(--text-primary)] break-words">{summary}</div>
+      ) : null}
+      {hasSummary && raw ? (
+        <button
+          type="button"
+          onClick={() => setShowRaw((value) => !value)}
+          className="mt-0.5 text-[var(--text-muted)] hover:text-cyan-400/90 underline underline-offset-2"
+        >
+          {showRaw ? 'hide raw' : 'show raw'}
+        </button>
+      ) : null}
+      {showRawBlock && raw ? (
+        <div className="mt-0.5 text-[var(--text-muted)] whitespace-pre-wrap break-words text-[9px] leading-relaxed max-h-24 overflow-y-auto">
+          {raw}
+        </div>
+      ) : null}
+    </div>
+  );
+}
+
 function formatDatalinkTime(value?: string): string {
   if (!value) return '--:--';
   try {
@@ -40,6 +78,8 @@ export default function DatalinkMessagesBlock({
   const [hint, setHint] = useState<string | null>(null);
   const [loadError, setLoadError] = useState<string | null>(null);
   const [priorityScanning, setPriorityScanning] = useState(false);
+  const [hiddenCount, setHiddenCount] = useState(0);
+  const [showHidden, setShowHidden] = useState(false);
   const pollUntilRef = useRef(0);
 
   const buildParams = useCallback(() => {
@@ -71,6 +111,7 @@ export default function DatalinkMessagesBlock({
         const json = await res.json();
         setConfigured(Boolean(json.configured));
         setMessages(Array.isArray(json.messages) ? json.messages : []);
+        setHiddenCount(typeof json.hidden_count === 'number' ? json.hidden_count : 0);
         setHint(typeof json.hint === 'string' ? json.hint : null);
         setLoadError(null);
         if (json.priority_scan || json.queued_refresh) {
@@ -163,16 +204,27 @@ export default function DatalinkMessagesBlock({
       ) : null}
       <div className="max-h-36 overflow-y-auto space-y-1.5 pr-1">
         {messages.map((message) => (
-          <div key={message.id} className="text-[10px] font-mono leading-snug border border-[var(--border-primary)]/60 bg-black/20 px-2 py-1.5">
-            <div className="flex items-center gap-2 text-[var(--text-muted)] mb-0.5">
-              <span>{formatDatalinkTime(message.timestamp)}</span>
-              {message.label ? <span className="text-orange-400/90">{message.label}</span> : null}
-              {message.source_type ? <span className="truncate">{message.source_type}</span> : null}
-            </div>
-            <div className="text-[var(--text-primary)] whitespace-pre-wrap break-words">{message.text}</div>
-          </div>
+          <DatalinkMessageRow key={message.id} message={message} />
         ))}
       </div>
+      {hiddenCount > 0 ? (
+        <p className="text-[9px] font-mono text-[var(--text-muted)] mt-1">
+          {hiddenCount} binary/fragment message{hiddenCount === 1 ? '' : 's'} hidden.{' '}
+          <button
+            type="button"
+            onClick={() => setShowHidden((value) => !value)}
+            className="underline underline-offset-2 hover:text-cyan-400/90"
+          >
+            {showHidden ? 'Hide note' : 'Why?'}
+          </button>
+          {showHidden ? (
+            <span className="block mt-0.5 text-[var(--text-muted)]/80">
+              VDL splits long telemetry into many frames. Southwest also uses proprietary formats
+              that cannot be decoded without airline keys.
+            </span>
+          ) : null}
+        </p>
+      ) : null}
     </div>
   );
 }