Fix: validate memory tenant model IDs on update and enforce tenant scope in memory pipeline (#14923)

### Related issues

Closes #14922

### What problem does this PR solve?

`POST /memories` already resolves `tenant_llm_id` and `tenant_embd_id`
through `ensure_tenant_model_id_for_params`, but `PUT
/memories/<memory_id>` accepted client-supplied `tenant_llm_id` /
`tenant_embd_id` without checking that those `tenant_llm` rows belong to
the memory owner’s tenant. A caller could persist another tenant’s row
IDs and later trigger extraction or embedding that loaded foreign model
credentials via `get_model_config_by_id(tenant_model_id)` with no tenant
allow-list.

This change aligns the update path with create: updates that change
models must go through `llm_id` / `embd_id` and
`ensure_tenant_model_id_for_params` scoped to the **memory’s**
`tenant_id` (not only the current user, so team-access cases stay
correct). Direct `tenant_*` fields in the body without `llm_id` /
`embd_id` are rejected. As defense in depth, `memory_message_service`
passes `allowed_tenant_ids` / `requester_tenant_id` into
`get_model_config_by_id` for LLM and embedding resolution so mismatched
IDs cannot be used even if bad data existed. A regression test rejects
payloads that set only `tenant_llm_id` / `tenant_embd_id`.

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

---------

Co-authored-by: jony376 <jony376@gmail.com>

api/db/joint_services/memory_message_service.py

@@ -154,7 +154,11 @@ async def extract_by_llm(tenant_id: str, tenant_llm_id: int, extract_conf: dict,
     else:
         user_prompts.append({"role": "user", "content": PromptAssembler.assemble_user_prompt(conversation_content, conversation_time, conversation_time)})
     if tenant_llm_id:
-        llm_config = get_model_config_by_id(tenant_llm_id)
+        llm_config = get_model_config_by_id(
+            tenant_llm_id,
+            allowed_tenant_ids=tenant_id,
+            requester_tenant_id=tenant_id,
+        )
     else:
         llm_config = get_model_config_by_type_and_name(tenant_id, LLMType.CHAT, llm_id)
     llm = LLMBundle(tenant_id, llm_config)
@@ -174,7 +178,11 @@ async def extract_by_llm(tenant_id: str, tenant_llm_id: int, extract_conf: dict,
 
 async def embed_and_save(memory, message_list: list[dict], task_id: str=None):
     if memory.tenant_embd_id:
-        embd_model_config = get_model_config_by_id(memory.tenant_embd_id)
+        embd_model_config = get_model_config_by_id(
+            memory.tenant_embd_id,
+            allowed_tenant_ids=memory.tenant_id,
+            requester_tenant_id=memory.tenant_id,
+        )
     else:
         embd_model_config = get_model_config_by_type_and_name(memory.tenant_id, LLMType.EMBEDDING, memory.embd_id)
     embedding_model = LLMBundle(memory.tenant_id, embd_model_config)
@@ -248,7 +256,11 @@ def query_message(filter_dict: dict, params: dict):
     question = question.strip()
     memory = memory_list[0]
     if memory.tenant_embd_id:
-        embd_model_config = get_model_config_by_id(memory.tenant_embd_id)
+        embd_model_config = get_model_config_by_id(
+            memory.tenant_embd_id,
+            allowed_tenant_ids=memory.tenant_id,
+            requester_tenant_id=memory.tenant_id,
+        )
     else:
         embd_model_config = get_model_config_by_type_and_name(memory.tenant_id, LLMType.EMBEDDING, memory.embd_id)
     embd_model = LLMBundle(memory.tenant_id, embd_model_config)