Dev 20260323 v2.0.11

.gitignore

@@ -234,6 +234,6 @@ apps/openwork-memos-integration/apps/desktop/public/assets/usecases/
 # Outputs and Evaluation Results
 outputs
 
-evaluation/data/temporal_locomo
+evaluation/data/
 test_add_pipeline.py
 test_file_pipeline.py

Makefile

@@ -36,7 +36,7 @@ pre_commit:
 	poetry run pre-commit run -a
 
 serve:
-	poetry run uvicorn memos.api.start_api:app
+	poetry run uvicorn memos.api.server_api:app
 
 openapi:
 	poetry run memos export_openapi --output docs/openapi.json

README.md

@@ -75,7 +75,7 @@
 - [**72% lower token usage**](https://x.com/MemOS_dev/status/2020854044583924111) — intelligent memory retrieval instead of loading full chat history
 - [**Multi-agent memory sharing**](https://x.com/MemOS_dev/status/2020538135487062094) — multi-instance agents share memory via same user_id, automatic context handoff
 
-Get your API key: [MemOS Dashboard](https://memos-dashboard.openmem.net/cn/login/)  
+Get your API key: [MemOS Dashboard](https://memos-dashboard.openmem.net/cn/login/)
 Full tutorial → [MemOS-Cloud-OpenClaw-Plugin](https://github.com/MemTensor/MemOS-Cloud-OpenClaw-Plugin)
 
 ### 🧠 Local Plugin — 100% On-Device Memory
@@ -84,7 +84,7 @@ Full tutorial → [MemOS-Cloud-OpenClaw-Plugin](https://github.com/MemTensor/Mem
 - **Hybrid search + task & skill evolution** — FTS5 + vector search, auto task summarization, reusable skills that self-upgrade
 - **Multi-agent collaboration + Memory Viewer** — memory isolation, skill sharing, full web dashboard with 7 management pages
 
- 🌐 [Homepage](https://memos-claw.openmem.net) · 
+ 🌐 [Homepage](https://memos-claw.openmem.net) ·
 📖 [Documentation](https://memos-claw.openmem.net/docs/index.html) · 📦 [NPM](https://www.npmjs.com/package/@memtensor/memos-local-openclaw-plugin)
 
 ## 📌 MemOS: Memory Operating System for AI Agents
@@ -104,10 +104,10 @@ Full tutorial → [MemOS-Cloud-OpenClaw-Plugin](https://github.com/MemTensor/Mem
 
 ### News
 
-- **2026-03-08** · 🦞 **MemOS OpenClaw Plugin — Cloud & Local**  
+- **2026-03-08** · 🦞 **MemOS OpenClaw Plugin — Cloud & Local**
   Official OpenClaw memory plugins launched. **Cloud Plugin**: hosted memory service with 72% lower token usage and multi-agent memory sharing ([MemOS-Cloud-OpenClaw-Plugin](https://github.com/MemTensor/MemOS-Cloud-OpenClaw-Plugin)). **Local Plugin** (`v1.0.0`): 100% on-device memory with persistent SQLite, hybrid search (FTS5 + vector), task summarization & skill evolution, multi-agent collaboration, and a full Memory Viewer dashboard.
 
-- **2025-12-24** · 🎉 **MemOS v2.0: Stardust (星尘) Release**  
+- **2025-12-24** · 🎉 **MemOS v2.0: Stardust (星尘) Release**
   Comprehensive KB (doc/URL parsing + cross-project sharing), memory feedback & precise deletion, multi-modal memory (images/charts), tool memory for agent planning, Redis Streams scheduling + DB optimizations, streaming/non-streaming chat, MCP upgrade, and lightweight quick/full deployment.
   <details>
     <summary>✨ <b>New Features</b></summary>
@@ -155,7 +155,7 @@ Full tutorial → [MemOS-Cloud-OpenClaw-Plugin](https://github.com/MemTensor/Mem
   </details>
 
 - **2025-08-07** · 🎉 **MemOS v1.0.0 (MemCube) Release**
-  First MemCube release with a word-game demo, LongMemEval evaluation, BochaAISearchRetriever integration, NebulaGraph support, improved search capabilities, and the official Playground launch.
+  First MemCube release with a word-game demo, LongMemEval evaluation, BochaAISearchRetriever integration, improved search capabilities, and the official Playground launch.
 
   <details>
     <summary>✨ <b>New Features</b></summary>
@@ -176,7 +176,7 @@ Full tutorial → [MemOS-Cloud-OpenClaw-Plugin](https://github.com/MemTensor/Mem
 
   **Plaintext Memory**
   - Integrated internet search with Bocha.
-  - Added support for Nebula database.
+  - Expanded graph database support.
   - Added contextual understanding for the tree-structured plaintext memory search interface.
 
   </details>
@@ -188,7 +188,7 @@ Full tutorial → [MemOS-Cloud-OpenClaw-Plugin](https://github.com/MemTensor/Mem
   - Fixed the concat_cache method.
 
   **Plaintext Memory**
-  - Fixed Nebula search-related issues.
+  - Fixed graph search-related issues.
 
   </details>
 
@@ -224,6 +224,7 @@ Full tutorial → [MemOS-Cloud-OpenClaw-Plugin](https://github.com/MemTensor/Mem
 2. Configure `docker/.env.example` and copy to `MemOS/.env`
  - The `OPENAI_API_KEY`,`MOS_EMBEDDER_API_KEY`,`MEMRADER_API_KEY` and others can be applied for through [`BaiLian`](https://bailian.console.aliyun.com/?spm=a2c4g.11186623.0.0.2f2165b08fRk4l&tab=api#/api).
  - Fill in the corresponding configuration in the `MemOS/.env` file.
+ - Supported LLM providers: **OpenAI**, **Azure OpenAI**, **Qwen (DashScope)**, **DeepSeek**, **MiniMax**, **Ollama**, **HuggingFace**, **vLLM**. Set `MOS_CHAT_MODEL_PROVIDER` to select the backend (e.g., `openai`, `qwen`, `deepseek`, `minimax`).
 3. Start the service.
 
 - Launch via Docker

apps/memos-local-openclaw/.env.example

@@ -18,6 +18,10 @@ SUMMARIZER_TEMPERATURE=0
 # Port for the web-based Memory Viewer (default: 18799)
 # VIEWER_PORT=18799
 
+# ─── Tavily Search (optional) ───
+# API key for Tavily web search (get from https://app.tavily.com)
+# TAVILY_API_KEY=tvly-your-tavily-api-key
+
 # ─── Telemetry (opt-out) ───
 # Anonymous usage analytics to help improve the plugin.
 # No memory content, queries, or personal data is ever sent — only tool names, latencies, and version info.

docker/.env.example

@@ -25,9 +25,12 @@ MOS_MAX_TOKENS=2048
 # Top-P for LLM in the Product API
 MOS_TOP_P=0.9
 # LLM for the Product API backend
-MOS_CHAT_MODEL_PROVIDER=openai             # openai | huggingface | vllm
+MOS_CHAT_MODEL_PROVIDER=openai             # openai | huggingface | vllm | minimax
 OPENAI_API_KEY=sk-xxx                      # [required] when provider=openai
 OPENAI_API_BASE=https://api.openai.com/v1  # [required] base for the key
+# MiniMax LLM (when provider=minimax)
+# MINIMAX_API_KEY=your-minimax-api-key     # [required] when provider=minimax
+# MINIMAX_API_BASE=https://api.minimax.io/v1  # base for MiniMax API
 
 ## MemReader / retrieval LLM
 MEMRADER_MODEL=gpt-4o-mini
@@ -80,8 +83,12 @@ EMBEDDING_MODEL=nomic-embed-text:latest
 ## Internet search & preference memory
 # Enable web search
 ENABLE_INTERNET=false
+# Internet search backend (bocha | tavily)
+INTERNET_SEARCH_BACKEND=bocha
 # API key for BOCHA Search
-BOCHA_API_KEY=                             # required if ENABLE_INTERNET=true
+BOCHA_API_KEY=                             # required if ENABLE_INTERNET=true and backend=bocha
+# API key for Tavily Search
+TAVILY_API_KEY=                            # required if ENABLE_INTERNET=true and backend=tavily
 # default search mode
 SEARCH_MODE=fast                          # fast | fine | mixture
 # Slow retrieval strategy configuration, rewrite is the rewrite strategy
@@ -127,7 +134,7 @@ MEMSCHEDULER_USE_REDIS_QUEUE=false
 
 ## Graph / vector stores
 # Neo4j database selection mode
-NEO4J_BACKEND=neo4j-community             # neo4j-community | neo4j | nebular | polardb
+NEO4J_BACKEND=neo4j-community             # neo4j-community | neo4j | polardb | postgres
 # Neo4j database url
 NEO4J_URI=bolt://localhost:7687           # required when backend=neo4j*
 #  Neo4j database user

docker/docker-compose.yml

@@ -26,7 +26,7 @@ services:
       - memos_network
 
   neo4j:
-    image: neo4j:5.26.4
+    image: neo4j:5.26.6
     container_name: neo4j-docker
     ports:
       - "7474:7474"   # HTTP

docker/requirements-full.txt

@@ -185,3 +185,4 @@ py-key-value-shared==0.2.8
 PyJWT==2.10.1
 pytest==9.0.2
 alibabacloud-oss-v2==1.2.2
+tavily-python==0.5.0

docker/requirements.txt

@@ -124,3 +124,4 @@ uvloop==0.22.1; sys_platform != 'win32'
 watchfiles==1.1.1
 websockets==15.0.1
 alibabacloud-oss-v2==1.2.2
+tavily-python==0.5.0

examples/basic_modules/llm.py

@@ -164,7 +164,37 @@
 print("Scenario 6:", resp)
 
 
-# Scenario 7: Using LLMFactory with Deepseek-chat + reasoning + CoT + streaming
+# Scenario 7: Using LLMFactory with MiniMax (OpenAI-compatible API)
+# Prerequisites:
+# 1. Get your API key from the MiniMax platform.
+# 2. Available models: MiniMax-M2.7 (flagship), MiniMax-M2.7-highspeed (low-latency),
+#    MiniMax-M2.5, MiniMax-M2.5-highspeed.
+
+cfg_mm = LLMConfigFactory.model_validate(
+    {
+        "backend": "minimax",
+        "config": {
+            "model_name_or_path": "MiniMax-M2.7",
+            "api_key": "your-minimax-api-key",
+            "api_base": "https://api.minimax.io/v1",
+            "temperature": 0.7,
+            "max_tokens": 1024,
+        },
+    }
+)
+llm = LLMFactory.from_config(cfg_mm)
+messages = [{"role": "user", "content": "Hello, who are you"}]
+resp = llm.generate(messages)
+print("Scenario 7:", resp)
+print("==" * 20)
+
+print("Scenario 7 (streaming):\n")
+for chunk in llm.generate_stream(messages):
+    print(chunk, end="")
+print("\n" + "==" * 20)
+
+
+# Scenario 8: Using LLMFactory with DeepSeek-chat + reasoning + CoT + streaming
 
 cfg2 = LLMConfigFactory.model_validate(
     {
@@ -186,7 +216,7 @@
         "content": "Explain how to solve this problem step-by-step. Be explicit in your thinking process. Question: If a train travels from city A to city B at 60 mph and returns at 40 mph, what is its average speed for the entire trip? Let's think step by step.",
     },
 ]
-print("Scenario 7:\n")
+print("Scenario 8:\n")
 for chunk in llm.generate_stream(messages):
     print(chunk, end="")
 print("==" * 20)

examples/basic_modules/neo4j_example.py

@@ -2,21 +2,36 @@
 
 from datetime import datetime
 
+from dotenv import load_dotenv
+
 from memos.configs.embedder import EmbedderConfigFactory
 from memos.configs.graph_db import GraphDBConfigFactory
 from memos.embedders.factory import EmbedderFactory
 from memos.graph_dbs.factory import GraphStoreFactory
 from memos.memories.textual.item import TextualMemoryItem, TreeNodeTextualMemoryMetadata
 
 
+load_dotenv()
+
+NEO4J_URI = os.getenv("NEO4J_URI", "bolt://localhost:7687")
+NEO4J_USER = os.getenv("NEO4J_USER", "neo4j")
+NEO4J_PASSWORD = os.getenv("NEO4J_PASSWORD", "12345678")
+NEO4J_DB_NAME = os.getenv("NEO4J_DB_NAME", "neo4j")
+EMBEDDING_DIMENSION = int(os.getenv("EMBEDDING_DIMENSION", "3072"))
+
+QDRANT_HOST = os.getenv("QDRANT_HOST", "localhost")
+QDRANT_PORT = int(os.getenv("QDRANT_PORT", "6333"))
+
 embedder_config = EmbedderConfigFactory.model_validate(
     {
-        "backend": "universal_api",
+        "backend": os.getenv("MOS_EMBEDDER_BACKEND", "universal_api"),
         "config": {
-            "provider": "openai",
-            "api_key": os.getenv("OPENAI_API_KEY", "sk-xxxxx"),
-            "model_name_or_path": "text-embedding-3-large",
-            "base_url": os.getenv("OPENAI_API_BASE", "https://api.openai.com/v1"),
+            "provider": os.getenv("MOS_EMBEDDER_PROVIDER", "openai"),
+            "api_key": os.getenv("MOS_EMBEDDER_API_KEY", os.getenv("OPENAI_API_KEY", "")),
+            "model_name_or_path": os.getenv("MOS_EMBEDDER_MODEL", "text-embedding-3-large"),
+            "base_url": os.getenv(
+                "MOS_EMBEDDER_API_BASE", os.getenv("OPENAI_API_BASE", "https://api.openai.com/v1")
+            ),
         },
     }
 )
@@ -31,12 +46,12 @@ def get_neo4j_graph(db_name: str = "paper"):
     config = GraphDBConfigFactory(
         backend="neo4j",
         config={
-            "uri": "bolt://xxxx:7687",
-            "user": "neo4j",
-            "password": "xxxx",
+            "uri": NEO4J_URI,
+            "user": NEO4J_USER,
+            "password": NEO4J_PASSWORD,
             "db_name": db_name,
             "auto_create": True,
-            "embedding_dimension": 3072,
+            "embedding_dimension": EMBEDDING_DIMENSION,
             "use_multi_db": True,
         },
     )
@@ -49,12 +64,12 @@ def example_multi_db(db_name: str = "paper"):
     config = GraphDBConfigFactory(
         backend="neo4j",
         config={
-            "uri": "bolt://localhost:7687",
-            "user": "neo4j",
-            "password": "12345678",
+            "uri": NEO4J_URI,
+            "user": NEO4J_USER,
+            "password": NEO4J_PASSWORD,
             "db_name": db_name,
             "auto_create": True,
-            "embedding_dimension": 3072,
+            "embedding_dimension": EMBEDDING_DIMENSION,
             "use_multi_db": True,
         },
     )
@@ -288,14 +303,14 @@ def example_shared_db(db_name: str = "shared-traval-group"):
         config = GraphDBConfigFactory(
             backend="neo4j",
             config={
-                "uri": "bolt://localhost:7687",
-                "user": "neo4j",
-                "password": "12345678",
+                "uri": NEO4J_URI,
+                "user": NEO4J_USER,
+                "password": NEO4J_PASSWORD,
                 "db_name": db_name,
                 "user_name": user_name,
                 "use_multi_db": False,
                 "auto_create": True,
-                "embedding_dimension": 3072,
+                "embedding_dimension": EMBEDDING_DIMENSION,
             },
         )
         # Step 2: Instantiate graph store
@@ -353,12 +368,12 @@ def example_shared_db(db_name: str = "shared-traval-group"):
     config_alice = GraphDBConfigFactory(
         backend="neo4j",
         config={
-            "uri": "bolt://localhost:7687",
-            "user": "neo4j",
-            "password": "12345678",
+            "uri": NEO4J_URI,
+            "user": NEO4J_USER,
+            "password": NEO4J_PASSWORD,
             "db_name": db_name,
             "user_name": user_list[0],
-            "embedding_dimension": 3072,
+            "embedding_dimension": EMBEDDING_DIMENSION,
         },
     )
     graph_alice = GraphStoreFactory.from_config(config_alice)
@@ -382,24 +397,22 @@ def run_user_session(
         config = GraphDBConfigFactory(
             backend="neo4j-community",
             config={
-                "uri": "bolt://localhost:7687",
-                "user": "neo4j",
-                "password": "12345678",
+                "uri": NEO4J_URI,
+                "user": NEO4J_USER,
+                "password": NEO4J_PASSWORD,
                 "db_name": db_name,
                 "user_name": user_name,
                 "use_multi_db": False,
-                "auto_create": False,  # Neo4j Community does not allow auto DB creation
-                "embedding_dimension": 3072,
+                "auto_create": False,
+                "embedding_dimension": EMBEDDING_DIMENSION,
                 "vec_config": {
-                    # Pass nested config to initialize external vector DB
-                    # If you use qdrant, please use Server instead of local mode.
                     "backend": "qdrant",
                     "config": {
                         "collection_name": "neo4j_vec_db",
-                        "vector_dimension": 3072,
+                        "vector_dimension": EMBEDDING_DIMENSION,
                         "distance_metric": "cosine",
-                        "host": "localhost",
-                        "port": 6333,
+                        "host": QDRANT_HOST,
+                        "port": QDRANT_PORT,
                     },
                 },
             },
@@ -408,14 +421,14 @@ def run_user_session(
         config = GraphDBConfigFactory(
             backend="neo4j",
             config={
-                "uri": "bolt://localhost:7687",
-                "user": "neo4j",
-                "password": "12345678",
+                "uri": NEO4J_URI,
+                "user": NEO4J_USER,
+                "password": NEO4J_PASSWORD,
                 "db_name": db_name,
                 "user_name": user_name,
                 "use_multi_db": False,
                 "auto_create": True,
-                "embedding_dimension": 3072,
+                "embedding_dimension": EMBEDDING_DIMENSION,
             },
         )
     graph = GraphStoreFactory.from_config(config)