Update

tests/unit/services/openai_gateway/gateway.py

@@ -107,6 +107,7 @@ def test_model_is_forced(local_db, monkeypatch):
             cfg,
             ("guest", "test"),
             "test",
+            "default",
         )
     )
     assert rt._client.calls[-1][1]["model"] == "deepseek-chat"
@@ -143,6 +144,7 @@ def test_model_route_overrides_upstream(local_db, monkeypatch):
                 cfg,
                 ("guest", "test"),
                 "test",
+                "default",
             )
         )
         url, body = rt._client.calls[-1]
@@ -169,7 +171,7 @@ def test_vision_rewrites_image_to_text(local_db, monkeypatch):
             ],
         }
     ]
-    run_async(rt.handle_chat({"messages": msgs}, cfg, ("guest", "test"), "test"))
+    run_async(rt.handle_chat({"messages": msgs}, cfg, ("guest", "test"), "test", "default"))
     sent = rt._client.calls[-1][1]["messages"][0]["content"]
     assert isinstance(sent, str) and "vision" in sent.lower()
 
@@ -185,6 +187,7 @@ def test_streaming_emits_sse(local_db, monkeypatch):
             cfg,
             ("guest", "test"),
             "test",
+            "default",
         )
     )
 
@@ -282,6 +285,7 @@ def test_embeddings_remaps_alias_to_configured_model(local_db, monkeypatch):
             cfg,
             ("guest", "test"),
             "test",
+            "default",
         )
     )
     assert rt._client.calls[-1][1]["model"] == "qwen/qwen3-embedding-8b"
@@ -302,6 +306,7 @@ def test_embeddings_success_records_ledger(local_db, monkeypatch):
             cfg,
             ("guest", "ledger_probe"),
             "test",
+            "default",
         )
     )
     assert resp.status_code == 200
@@ -325,7 +330,7 @@ def test_embeddings_disabled_returns_503(local_db, monkeypatch):
     rt = svc.runtime()
     resp = run_async(
         rt.handle_embeddings(
-            {"input": "hello"}, cfg, ("guest", "test"), "test"
+            {"input": "hello"}, cfg, ("guest", "test"), "test", "default"
         )
     )
     assert resp.status_code == 503
@@ -372,6 +377,7 @@ def test_images_remaps_alias_to_configured_model(local_db, monkeypatch):
             cfg,
             ("guest", "test"),
             "test",
+            "default",
         )
     )
     assert rt._client.calls[-1][1]["model"] == "black-forest-labs/flux.2-pro"
@@ -411,6 +417,7 @@ def test_image_route_overrides_upstream(local_db, monkeypatch):
                 cfg,
                 ("guest", "test"),
                 "test",
+                "default",
             )
         )
         url, body = rt._client.calls[-1]
@@ -435,6 +442,7 @@ def test_images_success_records_ledger(local_db, monkeypatch):
             cfg,
             ("guest", "img_ledger"),
             "test",
+            "default",
         )
     )
     assert resp.status_code == 200
@@ -459,7 +467,7 @@ def test_images_disabled_returns_503(local_db, monkeypatch):
     rt = svc.runtime()
     resp = run_async(
         rt.handle_images(
-            {"prompt": "badge"}, cfg, ("guest", "test"), "test"
+            {"prompt": "badge"}, cfg, ("guest", "test"), "test", "default"
         )
     )
     assert resp.status_code == 503
@@ -489,3 +497,271 @@ def test_compute_cost_embed_branch():
     assert output_cost == 0.0
     assert abs(total - 0.01) < 1e-9
     assert abs(input_cost - 0.01) < 1e-9
+
+
+def test_app_reference_stored_in_ledger_for_chat(local_db, monkeypatch):
+    monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeClient_openai_gateway)
+    svc = GatewayService()
+    cfg = svc.effective_config()
+    cfg["gateway_force_model"] = True
+    cfg["gateway_model"] = "deepseek-chat"
+    rt = svc.runtime()
+    run_async(
+        rt.handle_chat(
+            {"messages": [{"role": "user", "content": "ref_test"}]},
+            cfg,
+            ("guest", "ref_probe"),
+            "test-ua",
+            "my-custom-app",
+        )
+    )
+    row = get_table("gateway_usage_ledger").find_one(owner_id="ref_probe")
+    assert row is not None
+    assert row["app_reference"] == "my-custom-app"
+    assert row["backend"] == "chat"
+
+
+def test_app_reference_defaults_when_not_passed(local_db, monkeypatch):
+    monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeClient_openai_gateway)
+    svc = GatewayService()
+    cfg = svc.effective_config()
+    cfg["gateway_force_model"] = True
+    cfg["gateway_model"] = "deepseek-chat"
+    rt = svc.runtime()
+    run_async(
+        rt.handle_chat(
+            {"messages": [{"role": "user", "content": "default_test"}]},
+            cfg,
+            ("guest", "default_probe"),
+            "test-ua",
+            "default",
+        )
+    )
+    row = get_table("gateway_usage_ledger").find_one(owner_id="default_probe")
+    assert row is not None
+    assert row["app_reference"] == "default"
+
+
+def test_app_reference_stored_in_ledger_for_embeddings(local_db, monkeypatch):
+    monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeEmbedClient_openai_gateway)
+    svc = GatewayService()
+    cfg = svc.effective_config()
+    cfg["gateway_embed_enabled"] = True
+    cfg["gateway_embed_model"] = "qwen/qwen3-embedding-8b"
+    rt = svc.runtime()
+    run_async(
+        rt.handle_embeddings(
+            {"input": "hello"},
+            cfg,
+            ("guest", "embed_ref"),
+            "test-ua",
+            "devplace-embed-test-v-1-0-0",
+        )
+    )
+    row = get_table("gateway_usage_ledger").find_one(owner_id="embed_ref")
+    assert row is not None
+    assert row["app_reference"] == "devplace-embed-test-v-1-0-0"
+    assert row["backend"] == "embed"
+
+
+def test_app_reference_stored_in_ledger_for_images(local_db, monkeypatch):
+    monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeImageClient_openai_gateway)
+    svc = GatewayService()
+    cfg = svc.effective_config()
+    cfg["gateway_image_enabled"] = True
+    cfg["gateway_image_model"] = "black-forest-labs/flux.2-pro"
+    rt = svc.runtime()
+    run_async(
+        rt.handle_images(
+            {"prompt": "ref badge"},
+            cfg,
+            ("guest", "img_ref"),
+            "test-ua",
+            "devplace-image-test-v-1-0-0",
+        )
+    )
+    row = get_table("gateway_usage_ledger").find_one(owner_id="img_ref")
+    assert row is not None
+    assert row["app_reference"] == "devplace-image-test-v-1-0-0"
+    assert row["backend"] == "image"
+
+
+def test_app_reference_column_exists(local_db):
+    db = get_table("gateway_usage_ledger").db
+    if "gateway_usage_ledger" not in db.tables:
+        return
+    rows = list(db.query("PRAGMA table_info('gateway_usage_ledger')"))
+    column_names = [r["name"] for r in rows]
+    assert "app_reference" in column_names
+
+
+def test_chat_unknown_model_falls_back_to_default(local_db, monkeypatch):
+    monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeClient_openai_gateway)
+    svc = GatewayService()
+    cfg = svc.effective_config()
+    cfg["gateway_force_model"] = False
+    cfg["gateway_model"] = "deepseek-v4-flash"
+    rt = svc.runtime()
+    run_async(
+        rt.handle_chat(
+            {"model": "nonexistent-model-v99", "messages": [{"role": "user", "content": "hi"}]},
+            cfg,
+            ("guest", "fallback_test_chat"),
+            "test",
+            "default",
+        )
+    )
+    assert rt._client.calls[-1][1]["model"] == "deepseek-v4-flash"
+
+
+def test_embeddings_unknown_model_falls_back_to_default(local_db, monkeypatch):
+    monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeEmbedClient_openai_gateway)
+    svc = GatewayService()
+    cfg = svc.effective_config()
+    cfg["gateway_force_model"] = False
+    cfg["gateway_embed_enabled"] = True
+    cfg["gateway_embed_model"] = "qwen/qwen3-embedding-8b"
+    rt = svc.runtime()
+    run_async(
+        rt.handle_embeddings(
+            {"model": "nonexistent-embed-model", "input": "hello"},
+            cfg,
+            ("guest", "fallback_test_embed"),
+            "test",
+            "default",
+        )
+    )
+    assert rt._client.calls[-1][1]["model"] == "qwen/qwen3-embedding-8b"
+
+
+def test_images_unknown_model_falls_back_to_default(local_db, monkeypatch):
+    monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeImageClient_openai_gateway)
+    svc = GatewayService()
+    cfg = svc.effective_config()
+    cfg["gateway_force_model"] = False
+    cfg["gateway_image_enabled"] = True
+    cfg["gateway_image_model"] = "black-forest-labs/flux.2-pro"
+    rt = svc.runtime()
+    run_async(
+        rt.handle_images(
+            {"model": "nonexistent-image-model", "prompt": "test badge"},
+            cfg,
+            ("guest", "fallback_test_img"),
+            "test",
+            "default",
+        )
+    )
+    assert rt._client.calls[-1][1]["model"] == "black-forest-labs/flux.2-pro"
+
+
+class FakeClientWithUsage_openai_gateway(FakeClient_openai_gateway):
+    async def send(self, request):
+        self.calls.append((request.url, request.json_body))
+        body = request.json_body or {}
+        return FakeResp_openai_gateway(
+            payload={
+                "id": "x",
+                "model": body.get("model"),
+                "choices": [{"message": {"content": "hi there"}}],
+                "usage": {"prompt_tokens": 7, "completion_tokens": 3, "total_tokens": 10},
+            }
+        )
+
+
+def test_stream_options_stripped_from_upstream(local_db, monkeypatch):
+    monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeClient_openai_gateway)
+    svc = GatewayService()
+    cfg = svc.effective_config()
+    rt = svc.runtime()
+    run_async(
+        rt.handle_chat(
+            {
+                "messages": [{"role": "user", "content": "hi"}],
+                "stream": True,
+                "stream_options": {"include_usage": True},
+            },
+            cfg,
+            ("guest", "test"),
+            "test",
+            "default",
+        )
+    )
+    upstream_payload = rt._client.calls[-1][1]
+    assert "stream_options" not in upstream_payload
+    assert upstream_payload["stream"] is False
+
+
+def test_include_usage_emits_usage_chunk(local_db, monkeypatch):
+    monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeClientWithUsage_openai_gateway)
+    svc = GatewayService()
+    cfg = svc.effective_config()
+    rt = svc.runtime()
+    resp = run_async(
+        rt.handle_chat(
+            {
+                "messages": [{"role": "user", "content": "hi"}],
+                "stream": True,
+                "stream_options": {"include_usage": True},
+            },
+            cfg,
+            ("guest", "test"),
+            "test",
+            "default",
+        )
+    )
+
+    async def drain():
+        out = []
+        async for chunk in resp.body_iterator:
+            out.append(chunk if isinstance(chunk, str) else chunk.decode())
+        return "".join(out)
+
+    body = run_async(drain())
+    assert '"usage"' in body
+    assert '"total_tokens": 10' in body
+    assert "[DONE]" in body
+
+
+def test_no_usage_chunk_without_include_usage(local_db, monkeypatch):
+    monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeClientWithUsage_openai_gateway)
+    svc = GatewayService()
+    cfg = svc.effective_config()
+    rt = svc.runtime()
+    resp = run_async(
+        rt.handle_chat(
+            {"messages": [{"role": "user", "content": "hi"}], "stream": True},
+            cfg,
+            ("guest", "test"),
+            "test",
+            "default",
+        )
+    )
+
+    async def drain():
+        out = []
+        async for chunk in resp.body_iterator:
+            out.append(chunk if isinstance(chunk, str) else chunk.decode())
+        return "".join(out)
+
+    body = run_async(drain())
+    assert '"usage"' not in body
+    assert "[DONE]" in body
+
+
+def test_models_endpoint_publishes_molodetz(local_db):
+    from devplacepy.services.openai_gateway import routing
+
+    routing.seed_default_deepseek_routes()
+    try:
+        svc = GatewayService()
+        resp = svc._models_response()
+        payload = json.loads(resp.body.decode())
+        ids = {m["id"] for m in payload["data"]}
+        assert payload["object"] == "list"
+        assert {"molodetz", "molodetz-pro"} <= ids
+        for model in payload["data"]:
+            assert model["object"] == "model"
+            assert model["owned_by"] == "molodetz"
+    finally:
+        routing.model_store.remove("molodetz")
+        routing.model_store.remove("molodetz-pro")

tests/unit/services/openai_gateway/routing.py

@@ -280,3 +280,19 @@ def test_seed_default_deepseek_routes_is_idempotent_and_preserves_customization(
                 context_window=1_048_576,
             )
         )
+
+
+def test_seed_publishes_molodetz_aliases(local_db):
+    r.seed_default_deepseek_routes()
+    try:
+        molodetz = r.model_store.resolve("molodetz", "chat")
+        molodetz_pro = r.model_store.resolve("molodetz-pro", "chat")
+        assert molodetz is not None
+        assert molodetz.target_model == "deepseek-v4-flash"
+        assert molodetz_pro is not None
+        assert molodetz_pro.target_model == "deepseek-v4-pro"
+        sources = {row["source_model"] for row in r.model_store.list()}
+        assert {"molodetz", "molodetz-pro"} <= sources
+    finally:
+        r.model_store.remove("molodetz")
+        r.model_store.remove("molodetz-pro")

tests/unit/services/openai_gateway/usage.py

@@ -219,3 +219,32 @@ def test_compute_cost_native_upstream_cost_unaffected_by_tiering():
     total, _, _, native = compute_cost({"cost": 0.05}, norm, tiered, "chat")
     assert native is True
     assert total == 0.05
+
+
+def test_validate_app_reference_accepts_valid_slugs():
+    from devplacepy.services.openai_gateway.service import _validate_app_reference
+
+    assert _validate_app_reference("devplace-devii-v-1-0-0") == "devplace-devii-v-1-0-0"
+    assert _validate_app_reference("my-app") == "my-app"
+    assert _validate_app_reference("a") == "a"
+    assert _validate_app_reference("abcdefghijklmnopqrstuvwxyz0123") == "abcdefghijklmnopqrstuvwxyz0123"
+    assert _validate_app_reference("test_app.release-2") == "test_app.release-2"
+
+
+def test_validate_app_reference_rejects_invalid():
+    from devplacepy.services.openai_gateway.service import _validate_app_reference
+
+    assert _validate_app_reference("") == "default"
+    assert _validate_app_reference("   ") == "default"
+    assert _validate_app_reference("hello world") == "default"
+    assert _validate_app_reference("name@domain") == "default"
+    assert _validate_app_reference("name/domain") == "default"
+    assert _validate_app_reference(None) == "default"
+    assert _validate_app_reference("abcdefghijklmnopqrstuvwxyz01234") == "default"
+
+
+def test_validate_app_reference_strips_whitespace():
+    from devplacepy.services.openai_gateway.service import _validate_app_reference
+
+    assert _validate_app_reference("  my-app  ") == "my-app"
+    assert _validate_app_reference("\tdevplace\t") == "devplace"