diff --git a/public/packet-filter.js b/public/packet-filter.js
index 3bc4b2a7..d2082910 100644
--- a/public/packet-filter.js
+++ b/public/packet-filter.js
@@ -78,6 +78,18 @@
           while (i < len && /[a-zA-Z]/.test(input[i])) i++;
           var unit = input.slice(unitStart, i);
           if (!DURATION_UNITS[unit]) {
+            // Issue #1800: when the user typed a bare hex token (e.g.
+            // `path 2f0b001247a047ca`), the number/duration tokenizer eats
+            // only the leading "2f". Extend the scan to capture any further
+            // [0-9a-fA-F] chars; if the combined slice is pure hex of
+            // length >= 4, emit a targeted error rather than the cryptic
+            // duration-unit message.
+            var probe = i;
+            while (probe < len && /[0-9a-fA-F]/.test(input[probe])) probe++;
+            var fullSlice = input.slice(start, probe);
+            if (/^[0-9a-fA-F]+$/.test(fullSlice) && fullSlice.length >= 4) {
+              return { tokens: null, error: "Hex value must be quoted: try 'field == \"<hex>\"' or use the starts_with/contains operator" };
+            }
             return { tokens: null, error: "Invalid duration unit '" + unit + "' at position " + unitStart + " (expected s/m/h/d/w)" };
           }
           tokens.push({ type: TK.DURATION, value: parseFloat(numStr) * DURATION_UNITS[unit], raw: numStr + unit });
@@ -274,9 +286,26 @@
       if (isNaN(ms2)) return null;
       return Math.max(0, (Date.now() - ms2) / 1000);
     }
-    if (field === 'path') {
+    if (field === 'path' || field === 'path_prefixes') {
       try { return JSON.parse(packet.path_json || '[]').join(' → '); } catch(e) { return ''; }
     }
+    // Issue #1800: routed_through matches against resolved_path full pubkeys.
+    // resolved_path may arrive as a JSON string OR as an already-parsed array
+    // (depends on which API surface produced the packet). Handles both.
+    if (field === 'routed_through') {
+      var rp = packet.resolved_path;
+      if (rp == null) return '';
+      var arr;
+      if (typeof rp === 'string') {
+        try { arr = JSON.parse(rp); } catch (e) { return ''; }
+      } else if (Array.isArray(rp)) {
+        arr = rp;
+      } else {
+        return '';
+      }
+      if (!Array.isArray(arr)) return '';
+      return arr.map(function(h) { return String(h || '').toLowerCase(); }).join(' ');
+    }
     if (field === 'payload_bytes') {
       return packet.raw_hex ? Math.max(0, packet.raw_hex.length / 2 - 2) : 0;
     }
@@ -444,7 +473,9 @@
     { name: 'observer_iata', desc: 'Observer IATA region code (e.g. SJC, SFO)' },
     { name: 'iata',          desc: 'Alias of observer_iata' },
     { name: 'observations',  desc: 'Number of observations of this packet' },
-    { name: 'path',          desc: 'Hop path (joined with arrows)' },
+    { name: 'path',          desc: 'Hop path as 1-byte prefixes joined (e.g. a3→7f). For pubkey search use routed_through.' },
+    { name: 'path_prefixes', desc: 'Alias of path (1-byte hop prefixes joined with arrows)' },
+    { name: 'routed_through', desc: 'Match packets routed through a node by pubkey (full or prefix)' },
     { name: 'payload_bytes', desc: 'Payload size in bytes (size - 2 header bytes)' },
     { name: 'payload_hex',   desc: 'Payload bytes as hex (raw without header)' },
     { name: 'time',          desc: 'Packet timestamp (epoch ms)' },
diff --git a/test-packet-filter.js b/test-packet-filter.js
index fa95ea06..80fe64b7 100644
--- a/test-packet-filter.js
+++ b/test-packet-filter.js
@@ -255,5 +255,48 @@ test('#1774: payload.srcHash listed in FIELDS suggestions', () => {
   assert(names.indexOf('payload.srcHash') !== -1, 'payload.srcHash missing from FIELDS');
 });
 
+// --- Issue #1800: routed_through field + path desc + hex lexer error ---
+const routedPkt = {
+  ...pkt,
+  resolved_path: '["2f0b001247a047cabcdef0123456789a","aabbccddeeff00112233445566778899"]',
+};
+const routedArrPkt = {
+  ...pkt,
+  resolved_path: ['2f0b001247a047cabcdef0123456789a', 'aabbccddeeff00112233445566778899'],
+};
+test('#1800: routed_through starts_with prefix matches', () => {
+  assert(PF.compile('routed_through starts_with "2f0b00"').filter(routedPkt));
+});
+test('#1800: routed_through starts_with prefix matches (array form)', () => {
+  assert(PF.compile('routed_through starts_with "2f0b00"').filter(routedArrPkt));
+});
+test('#1800: routed_through == full pubkey matches', () => {
+  assert(PF.compile('routed_through contains "2f0b001247a047cabcdef0123456789a"').filter(routedPkt));
+});
+test('#1800: routed_through contains "2f0b" matches', () => {
+  assert(PF.compile('routed_through contains "2f0b"').filter(routedPkt));
+});
+test('#1800: routed_through does not match unrelated prefix', () => {
+  assert(!PF.compile('routed_through starts_with "deadbe"').filter(routedPkt));
+});
+test('#1800: bare hex value after path → "Hex value must be quoted" error', () => {
+  const c = PF.compile('path 2f0b001247a047ca');
+  assert(c.error !== null, 'should have parse/lex error');
+  assert(c.error.indexOf('Hex value must be quoted') !== -1,
+    'error must mention Hex value must be quoted, got: ' + c.error);
+});
+test('#1800 regression: path contains "a3" still matches existing prefix list', () => {
+  const prefixPkt = { ...pkt, path_json: '["a3","7f"]' };
+  assert(PF.compile('path contains "a3"').filter(prefixPkt));
+});
+test('#1800: routed_through listed in FIELDS suggestions', () => {
+  const names = PF.FIELDS.map(f => f.name);
+  assert(names.indexOf('routed_through') !== -1, 'routed_through missing from FIELDS');
+});
+test('#1800: path_prefixes alias resolves like path', () => {
+  const prefixPkt = { ...pkt, path_json: '["a3","7f"]' };
+  assert(PF.compile('path_prefixes contains "a3"').filter(prefixPkt));
+});
+
 console.log(`\n=== Results: ${pass} passed, ${fail} failed ===`);
 process.exit(fail > 0 ? 1 : 0);