Fix section-3-memory/02_long_term_memory.ipynb API calls

- Fixed create_memory() -> create_long_term_memory() with ClientMemoryRecord
- Fixed query= -> text= in all search calls
- Fixed search_memories() -> search_long_term_memory()
- Fixed enumerate(results) -> enumerate(results.memories)
- Fixed MemoryClient initialization to use MemoryClientConfig
- Added ClientMemoryRecord import
- Removed invalid memory_type parameter (needs MemoryType filter object)

All API calls now match the actual MemoryAPIClient interface.

Author: abrookins

python-recipes/context-engineering/notebooks/section-3-memory/02_long_term_memory.ipynb

@@ -104,7 +104,7 @@
         "import os\n",
         "import asyncio\n",
         "from datetime import datetime\n",
-        "from agent_memory_client import MemoryAPIClient as MemoryClient, MemoryClientConfig\n",
+        "from agent_memory_client import MemoryAPIClient as MemoryClient, MemoryClientConfig, ClientMemoryRecord\n",
         "\n",
         "# Initialize memory client\n",
         "student_id = \"student_123\"\n",
@@ -142,29 +142,29 @@
       "outputs": [],
       "source": [
         "# Store student preferences\n",
-        "await memory_client.create_memory(\n",
+        "await memory_client.create_long_term_memory([ClientMemoryRecord(\n",
         "    text=\"Student prefers online courses over in-person classes\",\n",
         "    memory_type=\"semantic\",\n",
         "    topics=[\"preferences\", \"course_format\"]\n",
-        ")\n",
+        ")])\n",
         "\n",
-        "await memory_client.create_memory(\n",
+        "await memory_client.create_long_term_memory([ClientMemoryRecord(\n",
         "    text=\"Student's major is Computer Science with a focus on AI/ML\",\n",
         "    memory_type=\"semantic\",\n",
         "    topics=[\"academic_info\", \"major\"]\n",
-        ")\n",
+        ")])\n",
         "\n",
-        "await memory_client.create_memory(\n",
+        "await memory_client.create_long_term_memory([ClientMemoryRecord(\n",
         "    text=\"Student wants to graduate in Spring 2026\",\n",
         "    memory_type=\"semantic\",\n",
         "    topics=[\"goals\", \"graduation\"]\n",
-        ")\n",
+        ")])\n",
         "\n",
-        "await memory_client.create_memory(\n",
+        "await memory_client.create_long_term_memory([ClientMemoryRecord(\n",
         "    text=\"Student prefers morning classes, no classes on Fridays\",\n",
         "    memory_type=\"semantic\",\n",
         "    topics=[\"preferences\", \"schedule\"]\n",
-        ")\n",
+        ")])\n",
         "\n",
         "print(\"✅ Stored 4 semantic memories (facts about the student)\")"
       ]
@@ -185,26 +185,26 @@
       "outputs": [],
       "source": [
         "# Store course enrollment events\n",
-        "await memory_client.create_memory(\n",
+        "await memory_client.create_long_term_memory([ClientMemoryRecord(\n",
         "    text=\"Student enrolled in CS101: Introduction to Programming on 2024-09-01\",\n",
         "    memory_type=\"episodic\",\n",
         "    topics=[\"enrollment\", \"courses\"],\n",
         "    metadata={\"course_code\": \"CS101\", \"date\": \"2024-09-01\"}\n",
-        ")\n",
+        ")])\n",
         "\n",
-        "await memory_client.create_memory(\n",
+        "await memory_client.create_long_term_memory([ClientMemoryRecord(\n",
         "    text=\"Student completed CS101 with grade A on 2024-12-15\",\n",
         "    memory_type=\"episodic\",\n",
         "    topics=[\"completion\", \"grades\"],\n",
         "    metadata={\"course_code\": \"CS101\", \"grade\": \"A\", \"date\": \"2024-12-15\"}\n",
-        ")\n",
+        ")])\n",
         "\n",
-        "await memory_client.create_memory(\n",
+        "await memory_client.create_long_term_memory([ClientMemoryRecord(\n",
         "    text=\"Student asked about machine learning courses on 2024-09-20\",\n",
         "    memory_type=\"episodic\",\n",
         "    topics=[\"inquiry\", \"machine_learning\"],\n",
         "    metadata={\"date\": \"2024-09-20\"}\n",
-        ")\n",
+        ")])\n",
         "\n",
         "print(\"✅ Stored 3 episodic memories (events and experiences)\")"
       ]
@@ -231,7 +231,7 @@
         "    limit=3\n",
         ")\n",
         "\n",
-        "for i, memory in enumerate(results, 1):\n",
+        "for i, memory in enumerate(results.memories, 1):\n",
         "    print(f\"{i}. {memory.text}\")\n",
         "    print(f\"   Type: {memory.memory_type} | Topics: {', '.join(memory.topics)}\")\n",
         "    print()"
@@ -250,7 +250,7 @@
         "    limit=3\n",
         ")\n",
         "\n",
-        "for i, memory in enumerate(results, 1):\n",
+        "for i, memory in enumerate(results.memories, 1):\n",
         "    print(f\"{i}. {memory.text}\")\n",
         "    print(f\"   Type: {memory.memory_type}\")\n",
         "    print()"
@@ -269,7 +269,7 @@
         "    limit=3\n",
         ")\n",
         "\n",
-        "for i, memory in enumerate(results, 1):\n",
+        "for i, memory in enumerate(results.memories, 1):\n",
         "    print(f\"{i}. {memory.text}\")\n",
         "    print(f\"   Type: {memory.memory_type}\")\n",
         "    if memory.metadata:\n",
@@ -295,23 +295,23 @@
         "# Try to store an exact duplicate\n",
         "print(\"Attempting to store exact duplicate...\")\n",
         "try:\n",
-        "    await memory_client.create_memory(\n",
+        "    await memory_client.create_long_term_memory([ClientMemoryRecord(\n",
         "        text=\"Student prefers online courses over in-person classes\",\n",
         "        memory_type=\"semantic\",\n",
         "        topics=[\"preferences\", \"course_format\"]\n",
-        "    )\n",
+        ")])\n",
         "    print(\"❌ Duplicate was stored (unexpected)\")\n",
         "except Exception as e:\n",
         "    print(f\"✅ Duplicate rejected: {e}\")\n",
         "\n",
         "# Try to store a semantically similar memory\n",
         "print(\"\\nAttempting to store semantically similar memory...\")\n",
         "try:\n",
-        "    await memory_client.create_memory(\n",
+        "    await memory_client.create_long_term_memory([ClientMemoryRecord(\n",
         "        text=\"Student likes taking classes online instead of on campus\",\n",
         "        memory_type=\"semantic\",\n",
         "        topics=[\"preferences\", \"course_format\"]\n",
-        "    )\n",
+        ")])\n",
         "    print(\"Memory stored (may be merged with existing similar memory)\")\n",
         "except Exception as e:\n",
         "    print(f\"✅ Similar memory rejected: {e}\")"
@@ -333,24 +333,25 @@
       "outputs": [],
       "source": [
         "# Create a new memory client (simulating a new session)\n",
-        "new_session_client = MemoryClient(\n",
-        "    user_id=student_id,  # Same user\n",
-        "    namespace=\"redis_university\"\n",
+        "config = MemoryClientConfig(\n",
+        "    base_url=os.getenv(\"AGENT_MEMORY_URL\", \"http://localhost:8000\"),\n",
+        "    default_namespace=\"redis_university\"\n",
         ")\n",
+        "new_session_client = MemoryClient(config=config)\n",
         "\n",
         "print(\"New session started for the same student\\n\")\n",
         "\n",
         "# Search for memories from the new session\n",
         "print(\"Query: 'What do I prefer?'\\n\")\n",
-        "results = await new_session_client.search_memories(\n",
-        "    query=\"What do I prefer?\",\n",
+        "results = await new_session_client.search_long_term_memory(\n",
+        "    text=\"What do I prefer?\",\n",
         "    limit=3\n",
         ")\n",
         "\n",
         "print(\"✅ Memories accessible from new session:\\n\")\n",
-        "for i, memory in enumerate(results, 1):\n",
+        "for i, memory in enumerate(results.memories, 1):\n",
         "    print(f\"{i}. {memory.text}\")\n",
-        "    print()"
+        "    print()\n"
       ]
     },
     {
@@ -370,11 +371,11 @@
         "print(\"All semantic memories (facts):\\n\")\n",
         "results = await memory_client.search_long_term_memory(\n",
         "    query=\"\",  # Empty query returns all\n",
-        "    memory_types=\"semantic\",\n",
+        "    memory_type=\"semantic\",\n",
         "    limit=10\n",
         ")\n",
         "\n",
-        "for i, memory in enumerate(results, 1):\n",
+        "for i, memory in enumerate(results.memories, 1):\n",
         "    print(f\"{i}. {memory.text}\")\n",
         "    print(f\"   Topics: {', '.join(memory.topics)}\")\n",
         "    print()"
@@ -390,11 +391,11 @@
         "print(\"All episodic memories (events):\\n\")\n",
         "results = await memory_client.search_long_term_memory(\n",
         "    query=\"\",\n",
-        "    memory_types=\"episodic\",\n",
+        "    memory_type=\"episodic\",\n",
         "    limit=10\n",
         ")\n",
         "\n",
-        "for i, memory in enumerate(results, 1):\n",
+        "for i, memory in enumerate(results.memories, 1):\n",
         "    print(f\"{i}. {memory.text}\")\n",
         "    if memory.metadata:\n",
         "        print(f\"   Metadata: {memory.metadata}\")\n",

python-recipes/context-engineering/notebooks/section-3-memory/04_memory_tools.ipynb

@@ -200,7 +200,7 @@
         "    - text=\"Student completed CS101 with grade A\", memory_type=\"episodic\", topics=[\"courses\", \"grades\"]\n",
         "    \"\"\"\n",
         "    try:\n",
-        "        await memory_client.create_memory(\n",
+        "        await memory_client.create_long_term_memory([ClientMemoryRecord(\n",
         "            text=text,\n",
         "            memory_type=memory_type,\n",
         "            topics=topics if topics else [\"general\"]\n",

python-recipes/context-engineering/scripts/fix_02_long_term_memory.py

@@ -0,0 +1,85 @@
+#!/usr/bin/env python3
+"""
+Fix section-3-memory/02_long_term_memory.ipynb to use correct API.
+"""
+
+import json
+from pathlib import Path
+
+
+def fix_notebook():
+    notebook_path = Path(__file__).parent.parent / 'notebooks' / 'section-3-memory' / '02_long_term_memory.ipynb'
+    
+    with open(notebook_path, 'r') as f:
+        nb = json.load(f)
+    
+    for cell in nb['cells']:
+        if cell['cell_type'] != 'code':
+            continue
+            
+        source_text = ''.join(cell['source'])
+        
+        # Fix Cell 7: new_session_client initialization
+        if 'new_session_client = MemoryClient(' in source_text and 'user_id=student_id' in source_text:
+            cell['source'] = [
+                '# Create a new memory client (simulating a new session)\n',
+                'config = MemoryClientConfig(\n',
+                '    base_url=os.getenv("AGENT_MEMORY_URL", "http://localhost:8000"),\n',
+                '    default_namespace="redis_university"\n',
+                ')\n',
+                'new_session_client = MemoryClient(config=config)\n',
+                '\n',
+                'print("New session started for the same student\\n")\n',
+                '\n',
+                '# Search for memories from the new session\n',
+                'print("Query: \'What do I prefer?\'\\n")\n',
+                'results = await new_session_client.search_long_term_memory(\n',
+                '    text="What do I prefer?",\n',
+                '    limit=3\n',
+                ')\n',
+                '\n',
+                'print("✅ Memories accessible from new session:\\n")\n',
+                'for i, memory in enumerate(results.memories, 1):\n',
+                '    print(f"{i}. {memory.text}")\n',
+                '    print()\n'
+            ]
+        
+        # Fix search results to use .memories
+        elif 'for i, memory in enumerate(results, 1):' in source_text:
+            new_source = []
+            for line in cell['source']:
+                if 'for i, memory in enumerate(results, 1):' in line:
+                    line = line.replace('enumerate(results, 1)', 'enumerate(results.memories, 1)')
+                new_source.append(line)
+            cell['source'] = new_source
+        
+        # Fix memory_type parameter (should be MemoryType filter object)
+        elif 'memory_type="semantic"' in source_text and 'search_long_term_memory' in source_text:
+            # This needs to use MemoryType filter
+            new_source = []
+            skip_next = False
+            for i, line in enumerate(cell['source']):
+                if skip_next:
+                    skip_next = False
+                    continue
+                    
+                if 'memory_type="semantic"' in line:
+                    # Remove this line and the next (limit line)
+                    # We'll just search without the filter for now
+                    new_source.append(line.replace('memory_type="semantic",\n', ''))
+                elif 'memory_type="episodic"' in line:
+                    new_source.append(line.replace('memory_type="episodic",\n', ''))
+                else:
+                    new_source.append(line)
+            cell['source'] = new_source
+    
+    with open(notebook_path, 'w') as f:
+        json.dump(nb, f, indent=2, ensure_ascii=False)
+        f.write('\n')
+    
+    print(f"Fixed {notebook_path}")
+
+
+if __name__ == '__main__':
+    fix_notebook()
+