diff --git a/internal/push/category.go b/internal/push/category.go
index 9ea14c3d..b48640dc 100644
--- a/internal/push/category.go
+++ b/internal/push/category.go
@@ -68,6 +68,8 @@ var categoryRegistry = []CategorySpec{
 		UnsupportedNote: "blocked on the ingestor's mask-sidecar support (data-ingestors#136)"},
 	{ID: "instance_segmentation", Family: FamilyImage, Label: "Instance segmentation", CLISupported: false,
 		UnsupportedNote: "not implemented"},
+	{ID: "causal_language_modeling", Family: FamilyText, Label: "Causal language modeling", CLISupported: false,
+		UnsupportedNote: "schema-recognized (data-ingestors#805); `tracebloc ingest` discover/build for its raw-.txt / prompt\\tcompletion `texts` layout is pending"},
 }
 
 // categoryByID indexes the registry for O(1) lookup, built once.
diff --git a/internal/push/category_registry_test.go b/internal/push/category_registry_test.go
index 1bb40ea3..997fc2ee 100644
--- a/internal/push/category_registry_test.go
+++ b/internal/push/category_registry_test.go
@@ -13,7 +13,7 @@ func TestRegistryKnownCategories(t *testing.T) {
 	want := []string{
 		"image_classification", "object_detection", "keypoint_detection",
 		"semantic_segmentation", "instance_segmentation",
-		"text_classification", "masked_language_modeling",
+		"text_classification", "masked_language_modeling", "causal_language_modeling",
 		"tabular_classification", "tabular_regression",
 		"time_series_forecasting", "time_to_event_prediction",
 	}
@@ -40,9 +40,9 @@ func TestSupportedCategories(t *testing.T) {
 			t.Errorf("SupportedCategoryIDs returned %q but IsCLISupported is false", id)
 		}
 	}
-	// semantic_/instance_segmentation are known but not yet pushable, and
-	// must explain why.
-	for _, id := range []string{"semantic_segmentation", "instance_segmentation"} {
+	// segmentation + causal_language_modeling are known but not yet pushable,
+	// and must explain why.
+	for _, id := range []string{"semantic_segmentation", "instance_segmentation", "causal_language_modeling"} {
 		if !IsKnown(id) {
 			t.Errorf("%s should be known", id)
 		}
diff --git a/internal/schema/ingest.v1.json b/internal/schema/ingest.v1.json
index 7f4f4a04..b5332b99 100644
--- a/internal/schema/ingest.v1.json
+++ b/internal/schema/ingest.v1.json
@@ -36,7 +36,8 @@
         "tabular_regression",
         "time_series_forecasting",
         "time_to_event_prediction",
-        "masked_language_modeling"
+        "masked_language_modeling",
+        "causal_language_modeling"
       ],
       "description": "Task category. Drives convention defaults: validators, data_format, default columns, default file extensions, default validator set."
     },
@@ -86,7 +87,7 @@
     "texts": {
       "type": "string",
       "minLength": 1,
-      "description": "Directory holding text files referenced by the labels CSV. Required for text_classification and token_classification."
+      "description": "Directory holding text files referenced by the labels CSV. Required for text_classification, token_classification, and causal_language_modeling (raw .txt: plain text, or a tab-separated prompt\\tcompletion pair)."
     },
 
     "sequences": {
@@ -336,6 +337,14 @@
       },
       "then": { "required": ["sequences"] }
     },
+    {
+      "description": "causal_language_modeling requires `texts` (raw .txt samples). It is self-supervised, so unlike text/token classification it does NOT require `label`.",
+      "if": {
+        "properties": { "category": { "const": "causal_language_modeling" } },
+        "required": ["category"]
+      },
+      "then": { "required": ["texts"] }
+    },
     {
       "description": "tabular and time-series categories require `schema`.",
       "if": {
@@ -411,7 +420,8 @@
         "properties": {
           "category": {
             "enum": [
-              "masked_language_modeling"
+              "masked_language_modeling",
+              "causal_language_modeling"
             ]
           }
         },