Clarify LLM control statement in automatic extraction section

Changed 'LLM has no control' to be more accurate:
- Your application's LLM can't directly control extraction
- But you can configure custom extraction prompts on the memory server
- The limitation is about client-side control, not configurability

Author: abrookins

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

@@ -45,13 +45,15 @@
         "\n",
         "**Pros:**\n",
         "- ✅ Fully automatic\n",
-        "- ✅ No LLM overhead\n",
+        "- ✅ No LLM overhead in your application\n",
         "- ✅ Consistent extraction\n",
         "\n",
         "**Cons:**\n",
-        "- ⚠️ LLM has no control\n",
+        "- ⚠️ Your application's LLM can't directly control what gets extracted\n",
         "- ⚠️ May extract too much or too little\n",
-        "- ⚠️ Can't decide what's important\n",
+        "- ⚠️ Can't dynamically decide what's important based on conversation context\n",
+        "\n",
+        "**Note:** You can configure custom extraction prompts on the memory server to guide what gets extracted, but your client application's LLM doesn't have direct control over the extraction process.\n",
         "\n",
         "#### 2. Tool-Based Memory (This Notebook)\n",
         "\n",