diff --git a/.github/workflows/tunnel-canary.yml b/.github/workflows/tunnel-canary.yml
index 9395250b0..360bea3f4 100644
--- a/.github/workflows/tunnel-canary.yml
+++ b/.github/workflows/tunnel-canary.yml
@@ -29,9 +29,10 @@ jobs:
           if [ -n "$err" ]; then
             echo "curl_stderr=$err"
           fi
-          # Ephemeral tunnels (pinggy-free, etc.) expire; DNS/connect 000 is inventory, not dual-gate.
-          if [ "${code:-000}" = "000" ] || [ "$curl_rc" -ne 0 ]; then
-            echo "::notice::tunnel unreachable (code=${code:-000} rc=$curl_rc); skip canary (not dual-gate)"
+          # Ephemeral tunnels expire. DNS/connect 000 is inventory, not dual-gate.
+          # A live host that does not advertise /health (404) is the same class.
+          if [ "${code:-000}" = "000" ] || [ "$curl_rc" -ne 0 ] || [ "$code" = "404" ]; then
+            echo "::notice::tunnel canary skipped (code=${code:-000} rc=$curl_rc); not dual-gate"
             exit 0
           fi
           [ "$code" = "200" ] || { echo "::error::tunnel health $code"; exit 1; }
diff --git a/archwiz/context_relationships/seed_merger.py b/archwiz/context_relationships/seed_merger.py
index 3009f5426..c935166ff 100755
--- a/archwiz/context_relationships/seed_merger.py
+++ b/archwiz/context_relationships/seed_merger.py
@@ -15,20 +15,28 @@
 except ImportError:  # Supports direct script use.
     from compiler import CompilationError, load_json
 
-MERGER_ID = "archwiz.context_relationships.seed_merger@1.2"
+MERGER_ID = "archwiz.context_relationships.seed_merger@1.3"
 
 # Line spans are observations of the same symbol, not identity. Collectors
 # disagree when one parse includes a decorator or trailing blank and another
-# does not. Missing observed_at must not fail the historical backfill.
+# does not. Historical seeds may store the same span as int or numeric string.
+# Missing observed_at must not fail the historical backfill.
 SPAN_ATTRIBUTE_KEYS = {"line", "start_line", "end_line"}
 
 
-def node_reference(record: Mapping[str, Any]) -> str:
-    kind = record.get("kind")
-    external_id = record.get("external_id")
-    if not isinstance(kind, str) or not isinstance(external_id, str):
-        raise CompilationError("seed node must contain string kind and external_id")
-    return f"{kind}:{external_id}"
+def span_line(value: Any) -> int | None:
+    """Coerce a span observation to int. Bool is not a line number."""
+    if isinstance(value, bool) or value is None:
+        return None
+    if isinstance(value, int):
+        return value
+    if isinstance(value, float) and value.is_integer():
+        return int(value)
+    if isinstance(value, str):
+        text = value.strip()
+        if text.isdigit() or (text.startswith("-") and text[1:].isdigit()):
+            return int(text)
+    return None
 
 
 def merge_attributes(
@@ -44,11 +52,18 @@ def merge_attributes(
     timestamps_available = isinstance(existing_observed_at, str) and isinstance(incoming_observed_at, str)
     for key, value in right.items():
         if key in merged and merged[key] != value:
-            if key in SPAN_ATTRIBUTE_KEYS and isinstance(merged[key], int) and isinstance(value, int):
-                if key == "end_line":
-                    merged[key] = max(merged[key], value)
-                else:
-                    merged[key] = min(merged[key], value)
+            if key in SPAN_ATTRIBUTE_KEYS:
+                left_line = span_line(merged[key])
+                right_line = span_line(value)
+                if left_line is None and right_line is None:
+                    continue
+                if left_line is None:
+                    merged[key] = right_line
+                    continue
+                if right_line is None:
+                    merged[key] = left_line
+                    continue
+                merged[key] = max(left_line, right_line) if key == "end_line" else min(left_line, right_line)
                 continue
             if not timestamps_available:
                 raise CompilationError(f"node {node_ref} has contradictory attribute {key!r}")
@@ -59,6 +74,14 @@ def merge_attributes(
     return merged
 
 
+def node_reference(record: Mapping[str, Any]) -> str:
+    kind = record.get("kind")
+    external_id = record.get("external_id")
+    if not isinstance(kind, str) or not isinstance(external_id, str):
+        raise CompilationError("seed node must contain string kind and external_id")
+    return f"{kind}:{external_id}"
+
+
 def canonical_github_blob_url(url: str, owner: str, name: str, ref: str) -> str:
     """Rewrite blob locators onto the seed ref. File identity is the path, not the blob SHA."""
     marker = f"https://github.com/{owner}/{name}/blob/"
diff --git a/tests/test_context_relationship_seed_merger.py b/tests/test_context_relationship_seed_merger.py
index 4cdeba7e1..15e4ea887 100644
--- a/tests/test_context_relationship_seed_merger.py
+++ b/tests/test_context_relationship_seed_merger.py
@@ -184,3 +184,33 @@ def test_merge_seeds_unions_symbol_spans_without_observed_at():
 
     assert merged["nodes"][0]["attributes"]["end_line"] == 120
     assert merged["nodes"][0]["attributes"]["line"] == 36
+
+
+def test_merge_seeds_coerces_string_end_line_without_observed_at():
+    first = {
+        **BASE,
+        "nodes": [
+            {
+                "kind": "symbol",
+                "external_id": "multi-ai-cli/backends/deepseek.py:DeepSeekBackend:37",
+                "attributes": {"name": "DeepSeekBackend", "line": 37, "end_line": 80},
+            }
+        ],
+        "edges": [],
+    }
+    second = {
+        **BASE,
+        "nodes": [
+            {
+                "kind": "symbol",
+                "external_id": "multi-ai-cli/backends/deepseek.py:DeepSeekBackend:37",
+                "attributes": {"name": "DeepSeekBackend", "line": "37", "end_line": "140"},
+            }
+        ],
+        "edges": [],
+    }
+
+    merged, _ = merge_seeds(first, second)
+
+    assert merged["nodes"][0]["attributes"]["end_line"] == 140
+    assert merged["nodes"][0]["attributes"]["line"] == 37