feat: enable multi-cube chat (read/write) & unify ChatRequest/ADDRequest normalization

docker/.env.example

@@ -1,60 +1,174 @@
-# MemOS Environment Variables Configuration
-TZ=Asia/Shanghai
+# MemOS Environment Variables (core runtime)
+# Legend: [required] needed for default startup; others are optional or conditional per comments.
 
-MOS_CUBE_PATH="/tmp/data_test"            # Path to memory storage (e.g. /tmp/data_test)
-MOS_ENABLE_DEFAULT_CUBE_CONFIG="true"     # Enable default cube config (true/false)
+## Base
+TZ=Asia/Shanghai
+ENV_NAME=PLAYGROUND_OFFLINE                 # Tag shown in DingTalk notifications (e.g., PROD_ONLINE/TEST); no runtime effect unless ENABLE_DINGDING_BOT=true
+MOS_CUBE_PATH=/tmp/data_test               # local data path
+MEMOS_BASE_PATH=.                          # CLI/SDK cache path
+MOS_ENABLE_DEFAULT_CUBE_CONFIG=true        # enable default cube config
+MOS_ENABLE_REORGANIZE=false                # enable memory reorg
+MOS_TEXT_MEM_TYPE=general_text             # general_text | tree_text
+ASYNC_MODE=sync                            # async/sync, used in default cube config
 
-# OpenAI Configuration
-OPENAI_API_KEY="sk-xxx"                   # Your OpenAI API key
-OPENAI_API_BASE="http://xxx"              # OpenAI API base URL (default: https://api.openai.com/v1)
+## User/session defaults
+MOS_USER_ID=root
+MOS_SESSION_ID=default_session
+MOS_MAX_TURNS_WINDOW=20
+MOS_TOP_K=50
 
-# MemOS Chat Model Configuration
+## Chat LLM (main dialogue)
 MOS_CHAT_MODEL=gpt-4o-mini
 MOS_CHAT_TEMPERATURE=0.8
 MOS_MAX_TOKENS=8000
 MOS_TOP_P=0.9
-MOS_TOP_K=50
-MOS_CHAT_MODEL_PROVIDER=openai
-
-# graph db
-# neo4j
-NEO4J_BACKEND=xxx
-NEO4J_URI=bolt://xxx
-NEO4J_USER=xxx
-NEO4J_PASSWORD=xxx
-MOS_NEO4J_SHARED_DB=xxx
-NEO4J_DB_NAME=xxx
-
-# tetxmem reog
-MOS_ENABLE_REORGANIZE=false
-
-# MemOS User Configuration
-MOS_USER_ID=root
-MOS_SESSION_ID=default_session
-MOS_MAX_TURNS_WINDOW=20
+MOS_CHAT_MODEL_PROVIDER=openai             # openai | huggingface | vllm
+MOS_MODEL_SCHEMA=memos.configs.llm.VLLMLLMConfig  # vllm only: config class path; keep default unless you extend it
+OPENAI_API_KEY=sk-xxx                      # [required] when provider=openai
+OPENAI_API_BASE=https://api.openai.com/v1  # [required] base for the key
+OPENAI_BASE_URL=                           # compatibility for eval/scheduler
+VLLM_API_KEY=                              # required when provider=vllm
+VLLM_API_BASE=http://localhost:8088/v1     # required when provider=vllm
 
-# MemRader Configuration
+## MemReader / retrieval LLM
 MEMRADER_MODEL=gpt-4o-mini
-MEMRADER_API_KEY=sk-xxx
-MEMRADER_API_BASE=http://xxx:3000/v1
+MEMRADER_API_KEY=sk-xxx                    # [required] can reuse OPENAI_API_KEY
+MEMRADER_API_BASE=http://localhost:3000/v1 # [required] base for the key
 MEMRADER_MAX_TOKENS=5000
 
-#embedding & rerank
+## Embedding & rerank
 EMBEDDING_DIMENSION=1024
-MOS_EMBEDDER_BACKEND=universal_api
-MOS_EMBEDDER_MODEL=bge-m3
-MOS_EMBEDDER_API_BASE=http://xxx
-MOS_EMBEDDER_API_KEY=EMPTY
-MOS_RERANKER_BACKEND=http_bge
-MOS_RERANKER_URL=http://xxx
-# Ollama Configuration (for embeddings)
-#OLLAMA_API_BASE=http://xxx
-
-# milvus for pref mem
-MILVUS_URI=http://xxx
-MILVUS_USER_NAME=xxx
-MILVUS_PASSWORD=xxx
-
-# pref mem
+MOS_EMBEDDER_BACKEND=universal_api        # universal_api | ollama
+MOS_EMBEDDER_PROVIDER=openai              # required when universal_api
+MOS_EMBEDDER_MODEL=bge-m3                 # siliconflow → use BAAI/bge-m3
+MOS_EMBEDDER_API_BASE=http://localhost:8000/v1 # required when universal_api
+MOS_EMBEDDER_API_KEY=EMPTY                 # required when universal_api
+OLLAMA_API_BASE=http://localhost:11434     # required when backend=ollama
+MOS_RERANKER_BACKEND=http_bge             # http_bge | http_bge_strategy | cosine_local
+MOS_RERANKER_URL=http://localhost:8001     # required when backend=http_bge*
+MOS_RERANKER_MODEL=bge-reranker-v2-m3     # siliconflow → use BAAI/bge-reranker-v2-m3
+MOS_RERANKER_HEADERS_EXTRA=               # extra headers, JSON string
+MOS_RERANKER_STRATEGY=single_turn
+MOS_RERANK_SOURCE=                        # optional rerank scope, e.g., history/stream/custom
+
+## Internet search & preference memory
+ENABLE_INTERNET=false
+BOCHA_API_KEY=                             # required if ENABLE_INTERNET=true
+SEARCH_MODE=fast                          # fast | fine | mixture
+FAST_GRAPH=false
+BM25_CALL=false
+VEC_COT_CALL=false
+FINE_STRATEGY=rewrite                     # rewrite | recreate | deep_search
+ENABLE_ACTIVATION_MEMORY=false
 ENABLE_PREFERENCE_MEMORY=true
-RETURN_ORIGINAL_PREF_MEM=true
+PREFERENCE_ADDER_MODE=fast                # fast | safe
+DEDUP_PREF_EXP_BY_TEXTUAL=false
+
+## Reader chunking
+MEM_READER_BACKEND=simple_struct           # simple_struct | strategy_struct
+MEM_READER_CHAT_CHUNK_TYPE=default         # default | content_length
+MEM_READER_CHAT_CHUNK_TOKEN_SIZE=1600      # tokens per chunk (default mode)
+MEM_READER_CHAT_CHUNK_SESS_SIZE=10         # sessions per chunk (default mode)
+MEM_READER_CHAT_CHUNK_OVERLAP=2            # overlap between chunks
+
+## Scheduler (MemScheduler / API)
+MOS_ENABLE_SCHEDULER=false
+MOS_SCHEDULER_TOP_K=10
+MOS_SCHEDULER_ACT_MEM_UPDATE_INTERVAL=300
+MOS_SCHEDULER_CONTEXT_WINDOW_SIZE=5
+MOS_SCHEDULER_THREAD_POOL_MAX_WORKERS=10000
+MOS_SCHEDULER_CONSUME_INTERVAL_SECONDS=0.01
+MOS_SCHEDULER_ENABLE_PARALLEL_DISPATCH=true
+MOS_SCHEDULER_ENABLE_ACTIVATION_MEMORY=false
+API_SCHEDULER_ON=true
+API_SEARCH_WINDOW_SIZE=5
+API_SEARCH_HISTORY_TURNS=5
+
+## Graph / vector stores
+NEO4J_BACKEND=neo4j-community             # neo4j-community | neo4j | nebular | polardb
+NEO4J_URI=bolt://localhost:7687           # required when backend=neo4j*
+NEO4J_USER=neo4j                          # required when backend=neo4j*
+NEO4J_PASSWORD=12345678                   # required when backend=neo4j*
+NEO4J_DB_NAME=neo4j                       # required for shared-db mode
+MOS_NEO4J_SHARED_DB=false
+QDRANT_HOST=localhost
+QDRANT_PORT=6333
+MILVUS_URI=http://localhost:19530         # required when ENABLE_PREFERENCE_MEMORY=true
+MILVUS_USER_NAME=root                     # same as above
+MILVUS_PASSWORD=12345678                  # same as above
+NEBULAR_HOSTS=["localhost"]
+NEBULAR_USER=root
+NEBULAR_PASSWORD=xxxxxx
+NEBULAR_SPACE=shared-tree-textual-memory
+NEBULAR_WORKING_MEMORY=20
+NEBULAR_LONGTERM_MEMORY=1000000
+NEBULAR_USER_MEMORY=1000000
+
+## Relational DB (user manager / PolarDB)
+MOS_USER_MANAGER_BACKEND=sqlite           # sqlite | mysql
+MYSQL_HOST=localhost                      # required when backend=mysql
+MYSQL_PORT=3306
+MYSQL_USERNAME=root
+MYSQL_PASSWORD=12345678
+MYSQL_DATABASE=memos_users
+MYSQL_CHARSET=utf8mb4
+POLAR_DB_HOST=localhost
+POLAR_DB_PORT=5432
+POLAR_DB_USER=root
+POLAR_DB_PASSWORD=123456
+POLAR_DB_DB_NAME=shared_memos_db
+POLAR_DB_USE_MULTI_DB=false
+
+## Redis (scheduler queue) — fill only if you want scheduler queues in Redis; otherwise in-memory queue is used
+REDIS_HOST=localhost                      # global Redis endpoint (preferred over MEMSCHEDULER_*)
+REDIS_PORT=6379
+REDIS_DB=0
+REDIS_PASSWORD=
+REDIS_SOCKET_TIMEOUT=
+REDIS_SOCKET_CONNECT_TIMEOUT=
+MEMSCHEDULER_REDIS_HOST=                  # fallback keys if not using the global ones
+MEMSCHEDULER_REDIS_PORT=
+MEMSCHEDULER_REDIS_DB=
+MEMSCHEDULER_REDIS_PASSWORD=
+MEMSCHEDULER_REDIS_TIMEOUT=
+MEMSCHEDULER_REDIS_CONNECT_TIMEOUT=
+
+## MemScheduler LLM
+MEMSCHEDULER_OPENAI_API_KEY=              # LLM key for scheduler’s own calls (OpenAI-compatible); leave empty if scheduler not using LLM
+MEMSCHEDULER_OPENAI_BASE_URL=             # Base URL for the above; can reuse OPENAI_API_BASE
+MEMSCHEDULER_OPENAI_DEFAULT_MODEL=gpt-4o-mini
+
+## Nacos (optional config center)
+NACOS_ENABLE_WATCH=false
+NACOS_WATCH_INTERVAL=60
+NACOS_SERVER_ADDR=
+NACOS_DATA_ID=
+NACOS_GROUP=DEFAULT_GROUP
+NACOS_NAMESPACE=
+AK=
+SK=
+
+## DingTalk bot & OSS upload
+ENABLE_DINGDING_BOT=false                 # set true -> fields below required
+DINGDING_ACCESS_TOKEN_USER=
+DINGDING_SECRET_USER=
+DINGDING_ACCESS_TOKEN_ERROR=
+DINGDING_SECRET_ERROR=
+DINGDING_ROBOT_CODE=
+DINGDING_APP_KEY=
+DINGDING_APP_SECRET=
+OSS_ENDPOINT=                             # bot image upload depends on OSS
+OSS_REGION=
+OSS_BUCKET_NAME=
+OSS_ACCESS_KEY_ID=
+OSS_ACCESS_KEY_SECRET=
+OSS_PUBLIC_BASE_URL=
+
+## Logging / external sink
+CUSTOM_LOGGER_URL=
+CUSTOM_LOGGER_TOKEN=
+CUSTOM_LOGGER_WORKERS=2
+
+## SDK / external client
+MEMOS_API_KEY=
+MEMOS_BASE_URL=https://memos.memtensor.cn/api/openmem/v1