Merge pull request #34 from redis/fix/fix-docker-compose

Fix docker-compose file

Author: abrookins

.github/workflows/python-tests.yml

@@ -34,19 +34,15 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        python-version: [3.12]  # Not testing with 3.13 at the moment
-        redis-version: ['6.2.6-v9', 'latest']  # 8.0-M03 is not working atm
+        python-version: [3.12]
+        redis-version: ['redis/redis-stack:6.2.6-v9', 'redis:8.0.3', 'redis:latest']
 
     steps:
     - uses: actions/checkout@v3
 
     - name: Set Redis image name
       run: |
-        if [[ "${{ matrix.redis-version }}" == "8.0-M03" ]]; then
-          echo "REDIS_IMAGE=redis:${{ matrix.redis-version }}" >> $GITHUB_ENV
-        else
-          echo "REDIS_IMAGE=redis/redis-stack-server:${{ matrix.redis-version }}" >> $GITHUB_ENV
-        fi
+          echo "REDIS_IMAGE=${{ matrix.redis-version }}" >> $GITHUB_ENV
 
     - name: Set up Python
       uses: actions/setup-python@v4
@@ -81,6 +77,12 @@ jobs:
     - name: Set up Docker Buildx
       uses: docker/setup-buildx-action@v3
 
+    - name: Log in to Docker Hub
+      uses: docker/login-action@v3
+      with:
+        username: ${{ secrets.DOCKER_USERNAME }}
+        password: ${{ secrets.DOCKER_TOKEN }}
+
     - name: Log in to GitHub Container Registry
       uses: docker/login-action@v3
       with:
@@ -103,6 +105,8 @@ jobs:
         platforms: linux/amd64,linux/arm64
         push: true
         tags: |
+          andrewbrookins510/agent-memory-server:latest
+          andrewbrookins510/agent-memory-server:${{ steps.version.outputs.version }}
           ghcr.io/${{ github.repository }}:latest
           ghcr.io/${{ github.repository }}:${{ steps.version.outputs.version }}
         cache-from: type=gha

agent-memory-client/agent_memory_client/__init__.py

@@ -5,7 +5,7 @@
 memory management capabilities for AI agents and applications.
 """
 
-__version__ = "0.9.0"
+__version__ = "0.9.1"
 
 from .client import MemoryAPIClient, MemoryClientConfig, create_memory_client
 from .exceptions import (

agent-memory-client/agent_memory_client/models.py

@@ -183,8 +183,8 @@ class WorkingMemory(BaseModel):
     )
 
     # TTL and timestamps
-    ttl_seconds: int = Field(
-        default=3600,  # 1 hour default
+    ttl_seconds: int | None = Field(
+        default=None,  # Persistent by default
         description="TTL for the working memory in seconds",
     )
     last_accessed: datetime = Field(

agent_memory_server/__init__.py

@@ -1,3 +1,3 @@
 """Redis Agent Memory Server - A memory system for conversational AI."""
 
-__version__ = "0.9.0"
+__version__ = "0.9.1"

agent_memory_server/mcp.py

@@ -137,7 +137,14 @@ async def run_sse_async(self):
 
         redis = await get_redis_conn()
         await ensure_search_index_exists(redis)
-        return await super().run_sse_async()
+
+        # Run the SSE server using our custom implementation
+        import uvicorn
+
+        app = self.sse_app()
+        await uvicorn.Server(
+            uvicorn.Config(app, host="0.0.0.0", port=int(self.settings.port))
+        ).serve()
 
     async def run_stdio_async(self):
         """Ensure Redis search index exists before starting STDIO MCP server."""

agent_memory_server/models.py

@@ -201,8 +201,8 @@ class WorkingMemory(BaseModel):
     )
 
     # TTL and timestamps
-    ttl_seconds: int = Field(
-        default=3600,  # 1 hour default
+    ttl_seconds: int | None = Field(
+        default=None,  # Persistent by default
         description="TTL for the working memory in seconds",
     )
     last_accessed: datetime = Field(

agent_memory_server/working_memory.py

@@ -121,7 +121,7 @@ async def get_working_memory(
             tokens=working_memory_data.get("tokens", 0),
             session_id=session_id,
             namespace=namespace,
-            ttl_seconds=working_memory_data.get("ttl_seconds", 3600),
+            ttl_seconds=working_memory_data.get("ttl_seconds", None),
             data=working_memory_data.get("data") or {},
             last_accessed=datetime.fromtimestamp(
                 working_memory_data.get("last_accessed", int(time.time())), UTC
@@ -188,18 +188,24 @@ async def set_working_memory(
     }
 
     try:
-        # Store with TTL
-        await redis_client.setex(
-            key,
-            working_memory.ttl_seconds,
-            json.dumps(
-                data, default=json_datetime_handler
-            ),  # Add custom handler for any remaining datetime objects
-        )
-        logger.info(
-            f"Set working memory for session {working_memory.session_id} with TTL {working_memory.ttl_seconds}s"
-        )
-
+        if working_memory.ttl_seconds is not None:
+            # Store with TTL
+            await redis_client.setex(
+                key,
+                working_memory.ttl_seconds,
+                json.dumps(data, default=json_datetime_handler),
+            )
+            logger.info(
+                f"Set working memory for session {working_memory.session_id} with TTL {working_memory.ttl_seconds}s"
+            )
+        else:
+            await redis_client.set(
+                key,
+                json.dumps(data, default=json_datetime_handler),
+            )
+            logger.info(
+                f"Set working memory for session {working_memory.session_id} with no TTL"
+            )
     except Exception as e:
         logger.error(
             f"Error setting working memory for session {working_memory.session_id}: {e}"

docker-compose.yml

@@ -34,7 +34,7 @@ services:
       dockerfile: Dockerfile
     environment:
       - REDIS_URL=redis://redis:6379
-      - PORT=9000
+      - PORT=9050
       # Add your API keys here or use a .env file
       - OPENAI_API_KEY=${OPENAI_API_KEY}
       - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}
@@ -46,19 +46,41 @@ services:
       - ENABLE_TOPIC_EXTRACTION=True
       - ENABLE_NER=True
     ports:
-      - "9000:9000"
+      - "9050:9000"
     depends_on:
       - redis
     command: ["uv", "run", "agent-memory", "mcp", "--mode", "sse"]
 
+  task-worker:
+    build:
+      context: .
+      dockerfile: Dockerfile
+    environment:
+      - REDIS_URL=redis://redis:6379
+      # Add your API keys here or use a .env file
+      - OPENAI_API_KEY=${OPENAI_API_KEY}
+      - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}
+      # Optional configurations with defaults
+      - LONG_TERM_MEMORY=True
+      - WINDOW_SIZE=20
+      - GENERATION_MODEL=gpt-4o-mini
+      - EMBEDDING_MODEL=text-embedding-3-small
+      - ENABLE_TOPIC_EXTRACTION=True
+      - ENABLE_NER=True
+    depends_on:
+      - redis
+    command: ["uv", "run", "agent-memory", "task-worker"]
+    volumes:
+      - ./agent_memory_server:/app/agent_memory_server
+    restart: unless-stopped
+
   redis:
-    image: redis/redis-stack:latest
+    image: redis:8
     ports:
-      - "16379:6379" # Redis port
-      - "18001:8001" # RedisInsight port
+      - "16380:6379" # Redis port
     volumes:
       - redis_data:/data
-    command: redis-stack-server --save 60 1 --loglevel warning
+    command: redis-server --save "" --loglevel warning --appendonly no --stop-writes-on-bgsave-error no
     healthcheck:
       test: [ "CMD", "redis-cli", "ping" ]
       interval: 30s

examples/memory_prompt_agent.py

@@ -34,9 +34,13 @@
 
 
 # Configure logging
-logging.basicConfig(level=logging.INFO)
+logging.basicConfig(level=logging.WARNING)
 logger = logging.getLogger(__name__)
 
+# Reduce third-party logging
+logging.getLogger("httpx").setLevel(logging.WARNING)
+logging.getLogger("openai").setLevel(logging.WARNING)
+
 # Environment setup
 MEMORY_SERVER_URL = os.getenv("MEMORY_SERVER_URL", "http://localhost:8000")
 DEFAULT_USER = "demo_user"
@@ -96,7 +100,6 @@ async def cleanup(self):
         """Clean up resources."""
         if self._memory_client:
             await self._memory_client.close()
-            logger.info("Memory client closed")
 
     async def _add_message_to_working_memory(
         self, session_id: str, user_id: str, role: str, content: str
@@ -145,8 +148,6 @@ async def _generate_response(
                 content = content["text"]
             messages.append({"role": msg["role"], "content": str(content)})
 
-        logger.info(f"Total messages for LLM: {len(messages)}")
-
         # Generate response
         response = self.llm.invoke(messages)
         return str(response.content)

examples/travel_agent.py

@@ -28,7 +28,6 @@
 import json
 import logging
 import os
-import textwrap
 
 from agent_memory_client import (
     MemoryAPIClient,
@@ -37,16 +36,26 @@
 from agent_memory_client.models import (
     WorkingMemory,
 )
-from langchain_community.tools.tavily_search import TavilySearchResults
 from langchain_core.callbacks.manager import CallbackManagerForToolRun
 from langchain_openai import ChatOpenAI
 from redis import Redis
 
 
+try:
+    from langchain_community.tools.tavily_search import TavilySearchResults
+except ImportError as e:
+    raise ImportError("Please install langchain-community for this demo.") from e
+
+
 # Configure logging
-logging.basicConfig(level=logging.INFO)
+logging.basicConfig(level=logging.WARNING)
 logger = logging.getLogger(__name__)
 
+# Reduce third-party logging
+logging.getLogger("httpx").setLevel(logging.WARNING)
+logging.getLogger("openai").setLevel(logging.WARNING)
+
+
 # Environment setup
 MEMORY_SERVER_URL = os.getenv("MEMORY_SERVER_URL", "http://localhost:8000")
 REDIS_URL = os.getenv("REDIS_URL", "redis://localhost:6379")
@@ -56,35 +65,35 @@
 
 SYSTEM_PROMPT = {
     "role": "system",
-    "content": textwrap.dedent("""
-                You are a helpful travel assistant. You can help with travel-related questions.
-                You have access to conversation history and memory management tools to provide
-                personalized responses.
-
-                Available tools:
-
-                1. **web_search** (if available): Search for current travel information, weather,
-                   events, or other up-to-date data when specifically needed.
-
-                2. **Memory Management Tools** (always available):
-                   - **search_memory**: Look up previous conversations and stored information
-                   - **get_working_memory**: Check current session context
-                   - **add_memory_to_working_memory**: Store important preferences or information
-                   - **update_working_memory_data**: Save session-specific data
-
-                **Guidelines**:
-                - Answer the user's actual question first and directly
-                - When someone shares information (like "I like X"), simply acknowledge it naturally - don't immediately give advice or suggestions unless they ask
-                - Search memory or web when it would be helpful for the current conversation
-                - Don't assume the user is actively planning a trip unless they explicitly say so
-                - Be conversational and natural - respond to what the user actually says
-                - When sharing memories, simply state what you remember rather than turning it into advice
-                - Only offer suggestions, recommendations, or tips if the user explicitly asks for them
-                - Store preferences and important details, but don't be overly eager about it
-                - If someone shares a preference, respond like a friend would - acknowledge it, maybe ask a follow-up question, but don't launch into advice
-
-                Be helpful, friendly, and responsive. Mirror their conversational style - if they're just chatting, chat back. If they ask for help, then help.
-                """),
+    "content": """
+    You are a helpful travel assistant. You can help with travel-related questions.
+    You have access to conversation history and memory management tools to provide
+    personalized responses.
+
+    Available tools:
+
+    1. **web_search** (if available): Search for current travel information, weather,
+       events, or other up-to-date data when specifically needed.
+
+    2. **Memory Management Tools** (always available):
+       - **search_memory**: Look up previous conversations and stored information
+       - **get_working_memory**: Check current session context
+       - **add_memory_to_working_memory**: Store important preferences or information
+       - **update_working_memory_data**: Save session-specific data
+
+    **Guidelines**:
+    - Answer the user's actual question first and directly
+    - When someone shares information (like "I like X"), simply acknowledge it naturally - don't immediately give advice or suggestions unless they ask
+    - Search memory or web when it would be helpful for the current conversation
+    - Don't assume the user is actively planning a trip unless they explicitly say so
+    - Be conversational and natural - respond to what the user actually says
+    - When sharing memories, simply state what you remember rather than turning it into advice
+    - Only offer suggestions, recommendations, or tips if the user explicitly asks for them
+    - Store preferences and important details, but don't be overly eager about it
+    - If someone shares a preference, respond like a friend would - acknowledge it, maybe ask a follow-up question, but don't launch into advice
+
+    Be helpful, friendly, and responsive. Mirror their conversational style - if they're just chatting, chat back. If they ask for help, then help.
+    """,
 }
 
 
@@ -151,12 +160,12 @@ def _setup_llms(self):
         # Define the web search tool function
         web_search_function = {
             "name": "web_search",
-            "description": textwrap.dedent("""
+            "description": """
               Search the web for current information about travel destinations,
               requirements, weather, events, or any other travel-related
               queries. Use this when you need up-to-date information that may
               not be in your training data.
-            """),
+            """,
             "parameters": {
                 "type": "object",
                 "properties": {