Merge branch 'dev' into main

Author: CaralHsi

README.md

@@ -43,7 +43,7 @@
 
 ---
 
-<img src="https://statics.memtensor.com.cn/memos/sota_score.jpg" alt="SOTA SCORE">
+<img src="https://cdn.memtensor.com.cn/img/1762436050812_3tgird_compressed.png" alt="SOTA SCORE">
 
 **MemOS** is an operating system for Large Language Models (LLMs) that enhances them with long-term memory capabilities. It allows LLMs to store, retrieve, and manage information, enabling more context-aware, consistent, and personalized interactions.
 
@@ -52,6 +52,28 @@
 - **API Reference**: https://memos-docs.openmem.net/docs/api/info/
 - **Source Code**: https://github.com/MemTensor/MemOS
 
+## 📰 News
+
+Stay up to date with the latest MemOS announcements, releases, and community highlights.
+
+
+- **2025-11-06** - 🎉 MemOS v1.1.3 (Async Memory & Preference):
+  Millisecond-level async memory add (support plain-text-memory and
+  preference memory); enhanced BM25, graph recall, and mixture search; full
+  results & code for LoCoMo, LongMemEval, PersonaMem, and PrefEval released.
+- **2025-10-30** - 🎉 MemOS v1.1.2 (API & MCP Update):
+API architecture overhaul and full MCP (Model Context Protocol) support — enabling models, IDEs, and agents to read/write external memory directly.
+- **2025-09-10** - 🎉 *MemOS v1.0.1 (Group Q&A Bot)*: Group Q&A bot based on MemOS Cube, updated KV-Cache performance comparison data across different GPU deployment schemes, optimized test benchmarks and statistics, added plaintext memory Reranker sorting, optimized plaintext memory hallucination issues, and Playground version updates. [Try PlayGround](https://memos-playground.openmem.net/login/)
+- **2025-08-07** - 🎉 *MemOS v1.0.0 (MemCube Release)*: First MemCube with word game demo, LongMemEval evaluation, BochaAISearchRetriever integration, NebulaGraph support, enhanced search capabilities, and official Playground launch.
+- **2025-07-29** – 🎉 *MemOS v0.2.2 (Nebula Update)*: Internet search+Nebula DB integration, refactored memory scheduler, KV Cache stress tests, MemCube Cookbook release (CN/EN), and 4b/1.7b/0.6b memory ops models.
+- **2025-07-21** – 🎉 *MemOS v0.2.1 (Neo Release)*: Lightweight Neo version with plaintext+KV Cache functionality, Docker/multi-tenant support, MCP expansion, and new Cookbook/Mud game examples.
+- **2025-07-11** – 🎉 *MemOS v0.2.0 (Cross-Platform)*: Added doc search/bilingual UI, MemReader-4B (local deploy), full Win/Mac/Linux support, and playground end-to-end connection.
+- **2025-07-07** – 🎉 *MemOS 1.0 (Stellar) Preview Release*: A SOTA Memory OS for LLMs is now open-sourced.
+- **2025-07-04** – 🎉 *MemOS Paper Released*: [MemOS: A Memory OS for AI System](https://arxiv.org/abs/2507.03724) was published on arXiv.
+- **2025-05-28** – 🎉 *Short Paper Uploaded*: [MemOS: An Operating System for Memory-Augmented Generation (MAG) in Large Language Models](https://arxiv.org/abs/2505.22101) was published on arXiv.
+- **2024-07-04** – 🎉 *Memory3 Model Released at WAIC 2024*: The new memory-layered architecture model was unveiled at the 2024 World Artificial Intelligence Conference.
+- **2024-07-01** – 🎉 *Memory3 Paper Released*: [Memory3: Language Modeling with Explicit Memory](https://arxiv.org/abs/2407.01178) introduces the new approach to structured memory in LLMs.
+
 ## 📈 Performance Benchmark
 
 MemOS demonstrates significant improvements over baseline memory solutions in multiple memory tasks,
@@ -265,19 +287,3 @@ We welcome contributions from the community! Please read our [contribution guide
 ## 📄 License
 
 MemOS is licensed under the [Apache 2.0 License](./LICENSE).
-
-## 📰 News
-
-Stay up to date with the latest MemOS announcements, releases, and community highlights.
-
-
-- **2025-09-10** - 🎉 *MemOS v1.0.1 (Group Q&A Bot)*: Group Q&A bot based on MemOS Cube, updated KV-Cache performance comparison data across different GPU deployment schemes, optimized test benchmarks and statistics, added plaintext memory Reranker sorting, optimized plaintext memory hallucination issues, and Playground version updates. [Try PlayGround](https://memos-playground.openmem.net/login/)
-- **2025-08-07** - 🎉 *MemOS v1.0.0 (MemCube Release)*: First MemCube with word game demo, LongMemEval evaluation, BochaAISearchRetriever integration, NebulaGraph support, enhanced search capabilities, and official Playground launch.
-- **2025-07-29** – 🎉 *MemOS v0.2.2 (Nebula Update)*: Internet search+Nebula DB integration, refactored memory scheduler, KV Cache stress tests, MemCube Cookbook release (CN/EN), and 4b/1.7b/0.6b memory ops models.
-- **2025-07-21** – 🎉 *MemOS v0.2.1 (Neo Release)*: Lightweight Neo version with plaintext+KV Cache functionality, Docker/multi-tenant support, MCP expansion, and new Cookbook/Mud game examples.
-- **2025-07-11** – 🎉 *MemOS v0.2.0 (Cross-Platform)*: Added doc search/bilingual UI, MemReader-4B (local deploy), full Win/Mac/Linux support, and playground end-to-end connection.
-- **2025-07-07** – 🎉 *MemOS 1.0 (Stellar) Preview Release*: A SOTA Memory OS for LLMs is now open-sourced.
-- **2025-07-04** – 🎉 *MemOS Paper Released*: [MemOS: A Memory OS for AI System](https://arxiv.org/abs/2507.03724) was published on arXiv.
-- **2025-05-28** – 🎉 *Short Paper Uploaded*: [MemOS: An Operating System for Memory-Augmented Generation (MAG) in Large Language Models](https://arxiv.org/abs/2505.22101) was published on arXiv.
-- **2024-07-04** – 🎉 *Memory3 Model Released at WAIC 2024*: The new memory-layered architecture model was unveiled at the 2024 World Artificial Intelligence Conference.
-- **2024-07-01** – 🎉 *Memory3 Paper Released*: [Memory3: Language Modeling with Explicit Memory](https://arxiv.org/abs/2407.01178) introduces the new approach to structured memory in LLMs.

evaluation/.env-example

@@ -21,4 +21,3 @@ MEMU_API_KEY="mu_xxx"
 SUPERMEMORY_API_KEY="sm_xxx"
 MEMOBASE_API_KEY="xxx"
 MEMOBASE_PROJECT_URL="http://***.***.***.***:8019"
-

examples/api/__init__.py

@@ -0,0 +1,144 @@
+#!/usr/bin/env python3
+"""
+Simulate full MemOS Product API workflow:
+1. Register user
+2. Add memory
+3. Search memory
+4. Chat (stream)
+"""
+
+import json
+
+import requests
+
+
+BASE_URL = "http://0.0.0.0:8001/product"
+HEADERS = {"Content-Type": "application/json"}
+
+index = "24"
+USER_ID = f"memos_user_id_{index}"
+USER_NAME = f"memos_user_alice_{index}"
+MEM_CUBE_ID = f"memos_cube_id_{index}"
+SESSION_ID = f"memos_session_id_{index}"
+SESSION_ID2 = f"memos_session_id_{index}_s2"
+
+
+def register_user():
+    url = f"{BASE_URL}/users/register"
+    data = {
+        "user_id": USER_ID,
+        "user_name": USER_NAME,
+        "interests": "memory,retrieval,test",
+        "mem_cube_id": MEM_CUBE_ID,
+    }
+    print(f"[*] Registering user {USER_ID} ...")
+    resp = requests.post(url, headers=HEADERS, data=json.dumps(data), timeout=30)
+    print(resp.status_code, resp.text)
+    return resp.json()
+
+
+def add_memory():
+    url = f"{BASE_URL}/add"
+    data = {
+        "user_id": USER_ID,
+        "memory_content": "今天我在测试 MemOS 的记忆添加与检索流程。",
+        "messages": [{"role": "user", "content": "我今天在做系统测试"}],
+        "doc_path": None,
+        "mem_cube_id": MEM_CUBE_ID,
+        "source": "test_script",
+        "user_profile": False,
+        "session_id": SESSION_ID,
+    }
+    print("[*] Adding memory ...")
+    resp = requests.post(url, headers=HEADERS, data=json.dumps(data), timeout=30)
+    print(resp.status_code, resp.text)
+    return resp.json()
+
+
+def search_memory(query="系统测试"):
+    url = f"{BASE_URL}/search"
+    data = {
+        "user_id": USER_ID,
+        "query": query,
+        "mem_cube_id": MEM_CUBE_ID,
+        "top_k": 5,
+        "session_id": SESSION_ID,
+    }
+    print("[*] Searching memory ...")
+    resp = requests.post(url, headers=HEADERS, data=json.dumps(data), timeout=30)
+    print(resp.status_code, resp.text)
+    return resp.json()
+
+
+def chat_stream(query: str, session_id: str, history: list | None = None):
+    url = f"{BASE_URL}/chat"
+    data = {
+        "user_id": USER_ID,
+        "query": query,
+        "mem_cube_id": MEM_CUBE_ID,
+        "history": history,
+        "internet_search": False,
+        "moscube": False,
+        "session_id": session_id,
+    }
+
+    print("[*] Starting streaming chat ...")
+
+    with requests.post(url, headers=HEADERS, data=json.dumps(data), stream=True) as resp:
+        for raw_line in resp.iter_lines():
+            if not raw_line:
+                continue
+            line = raw_line.decode("utf-8", errors="ignore")
+
+            payload = line.removeprefix("data: ").strip()
+            if payload == "[DONE]":
+                print("[done]")
+                break
+
+            try:
+                msg = json.loads(payload)
+                msg_type = msg.get("type")
+                msg_data = msg.get("data") or msg.get("content")
+
+                if msg_type == "text":
+                    print(msg_data, end="", flush=True)
+                elif msg_type == "reference":
+                    print(f"\n[参考记忆] {msg_data}")
+                elif msg_type == "status":
+                    pass
+                elif msg_type == "suggestion":
+                    print(f"\n[建议] {msg_data}")
+                elif msg_type == "end":
+                    print("\n[✅ Chat End]")
+                else:
+                    print(f"\n[{msg_type}] {msg_data}")
+            except Exception:
+                try:
+                    print(payload.encode("latin-1").decode("utf-8"), end="")
+                except Exception:
+                    print(payload)
+
+
+if __name__ == "__main__":
+    print("===== STEP 1: Register User =====")
+    register_user()
+
+    print("\n===== STEP 2: Add Memory =====")
+    add_memory()
+
+    print("\n===== STEP 3: Search Memory =====")
+    search_memory()
+
+    print("\n===== STEP 4: Stream Chat =====")
+    chat_stream("我很开心，我今天吃了好吃的拉面", SESSION_ID, history=[])
+    chat_stream(
+        "我刚和你说什么",
+        SESSION_ID,
+        history=[
+            {"role": "user", "content": "我很开心，我今天吃了好吃的拉面"},
+            {"role": "assistant", "content": "🉑"},
+        ],
+    )
+
+    print("\n===== STEP 4: Stream Chat =====")
+    chat_stream("我刚和你说什么了呢", SESSION_ID2, history=[])

examples/api/product_api.py

@@ -0,0 +1,62 @@
+from memos.api.routers.server_router import mem_scheduler
+from memos.mem_scheduler.schemas.message_schemas import ScheduleMessageItem
+
+
+# Debug: Print scheduler configuration
+print("=== Scheduler Configuration Debug ===")
+print(f"Scheduler type: {type(mem_scheduler).__name__}")
+print(f"Config: {mem_scheduler.config}")
+print(f"use_redis_queue: {mem_scheduler.use_redis_queue}")
+print(f"Queue type: {type(mem_scheduler.memos_message_queue).__name__}")
+print(f"Queue maxsize: {getattr(mem_scheduler.memos_message_queue, 'maxsize', 'N/A')}")
+
+# Check if Redis queue is connected
+if hasattr(mem_scheduler.memos_message_queue, "_is_connected"):
+    print(f"Redis connected: {mem_scheduler.memos_message_queue._is_connected}")
+if hasattr(mem_scheduler.memos_message_queue, "_redis_conn"):
+    print(f"Redis connection: {mem_scheduler.memos_message_queue._redis_conn}")
+print("=====================================\n")
+
+queue = mem_scheduler.memos_message_queue
+queue.clear()
+
+
+# 1. Define a handler function
+def my_test_handler(messages: list[ScheduleMessageItem]):
+    print(f"My test handler received {len(messages)} messages:")
+    for msg in messages:
+        print(f" my_test_handler - {msg.item_id}: {msg.content}")
+        print(
+            f"{queue._redis_conn.xinfo_groups(queue.stream_name)} qsize: {queue.qsize()} messages:{messages}"
+        )
+
+
+# 2. Register the handler
+TEST_HANDLER_LABEL = "test_handler"
+mem_scheduler.register_handlers({TEST_HANDLER_LABEL: my_test_handler})
+
+# 3. Create messages
+messages_to_send = [
+    ScheduleMessageItem(
+        item_id=f"test_item_{i}",
+        user_id="test_user",
+        mem_cube_id="test_mem_cube",
+        label=TEST_HANDLER_LABEL,
+        content=f"This is test message {i}",
+    )
+    for i in range(5)
+]
+
+# 5. Submit messages
+for mes in messages_to_send:
+    print(f"Submitting message {mes.item_id} to the scheduler...")
+    mem_scheduler.submit_messages([mes])
+
+# 6. Wait for messages to be processed (limited to 100 checks)
+print("Waiting for messages to be consumed (max 100 checks)...")
+mem_scheduler.mem_scheduler_wait()
+
+
+# 7. Stop the scheduler
+print("Stopping the scheduler...")
+mem_scheduler.stop()