Document voice_ask working keywords from testing

Tested AI Chat API with different queries:

WORKING (return real patient data):
- "how is dad" / "how is mom" → full status report
- "bathroom" → restroom usage times
- "kitchen" → cooking activity
- "sleep" / "night" → sleep hours, wake ups
- "shower" → shower history
- "temperature" → room temperature (verified!)
- "last seen" → online status

NOT WORKING (generic AI responses):
- status, summary, location, daily report

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

wellnuoSheme/01_ENV_Credentials and existing API.json

@@ -64,7 +64,7 @@
         "backend",
         "verified"
       ],
-      "description": "function=voice_ask\n\nRequest:\n- clientId: string (e.g. \"MA_001\")\n- user_name: string\n- token: string (JWT from auth)\n- question: string (e.g. \"how is dad doing\")\n- deployment_id: string (patient ID)\n\nResponse:\n{\n  \"ok\": true,\n  \"response\": {\n    \"Command\": string,\n    \"body\": string (AI response text),\n    \"language\": string\n  },\n  \"status\": \"200 OK\"\n}\n\nUsed for: AI assistant chat about patient status",
+      "description": "function=voice_ask\n\nRequest:\n- clientId: string (e.g. \"MA_001\")\n- user_name: string\n- token: string (JWT from auth)\n- question: string (keyword-based!)\n- deployment_id: string (patient ID)\n\nWORKING KEYWORDS:\n- \"how is dad\" / \"how is mom\" → full status\n- \"bathroom\" → restroom usage times\n- \"kitchen\" → cooking activity\n- \"sleep\" / \"night\" → sleep hours, wake ups\n- \"shower\" → shower history\n- \"temperature\" → room temp (REAL DATA!)\n- \"last seen\" → online status\n\nNOT WORKING (generic AI):\n- status, summary, location, daily report\n\nNote: Only specific keywords trigger real patient data context!",
       "x": 520,
       "y": 400
     },