diff --git a/bot/commands/medications.py b/bot/commands/medications.py
index 3386a42..003b674 100644
--- a/bot/commands/medications.py
+++ b/bot/commands/medications.py
@@ -3,10 +3,46 @@ Medications command handler - bot-side hooks for medication management
 """
 
 import asyncio
+import re
 from bot.command_registry import register_module
 import ai.parser as ai_parser
 
 
+async def _get_scheduled_time_from_context(message, med_name):
+    """Fetch recent messages and extract scheduled time from medication reminder.
+
+    Looks for bot messages in the last 5 messages that match the pattern:
+    "Time to take {med_name} (...) · HH:MM"
+
+    Returns the scheduled time string (e.g., "12:00") or None if not found.
+    """
+    try:
+        # Get last 5 messages from channel history
+        async for msg in message.channel.history(limit=5):
+            # Skip the current message
+            if msg.id == message.id:
+                continue
+
+            # Check if this is a bot message
+            if not msg.author.bot:
+                continue
+
+            content = msg.content
+
+            # Look for reminder pattern: "Time to take {med_name} (...) · HH:MM"
+            # Use case-insensitive match for medication name
+            pattern = rf"Time to take {re.escape(med_name)}.*?·\s*(\d{{1,2}}:\d{{2}})"
+            match = re.search(pattern, content, re.IGNORECASE)
+
+            if match:
+                return match.group(1)
+
+    except Exception:
+        pass
+
+    return None
+
+
 async def handle_medication(message, session, parsed):
     action = parsed.get("action", "unknown")
     token = session["token"]
@@ -19,10 +55,15 @@ async def handle_medication(message, session, parsed):
             if not meds:
                 await message.channel.send("You don't have any medications yet.")
             else:
-                lines = [f"- **{m['name']}**: {m['dosage']} {m['unit']} ({m.get('frequency', 'n/a')})" for m in meds]
+                lines = [
+                    f"- **{m['name']}**: {m['dosage']} {m['unit']} ({m.get('frequency', 'n/a')})"
+                    for m in meds
+                ]
                 await message.channel.send("**Your medications:**\n" + "\n".join(lines))
         else:
-            await message.channel.send(f"Error: {resp.get('error', 'Failed to fetch medications')}")
+            await message.channel.send(
+                f"Error: {resp.get('error', 'Failed to fetch medications')}"
+            )
 
     elif action == "add":
         name = parsed.get("name")
@@ -44,7 +85,7 @@ async def handle_medication(message, session, parsed):
             # Store pending action in session for confirmation
             if "pending_confirmations" not in session:
                 session["pending_confirmations"] = {}
-            
+
             confirmation_id = f"med_add_{name}"
             session["pending_confirmations"][confirmation_id] = {
                 "action": "add",
@@ -54,10 +95,12 @@ async def handle_medication(message, session, parsed):
                 "frequency": frequency,
                 "times": times,
                 "days_of_week": days_of_week,
-                "interval_days": interval_days
+                "interval_days": interval_days,
             }
-            
-            schedule_desc = _format_schedule(frequency, times, days_of_week, interval_days)
+
+            schedule_desc = _format_schedule(
+                frequency, times, days_of_week, interval_days
+            )
             await message.channel.send(
                 f"{confirmation_prompt}\n\n"
                 f"**Details:**\n"
@@ -68,23 +111,52 @@ async def handle_medication(message, session, parsed):
             )
             return
 
-        await _add_medication(message, token, name, dosage, unit, frequency, times, days_of_week, interval_days)
+        await _add_medication(
+            message,
+            token,
+            name,
+            dosage,
+            unit,
+            frequency,
+            times,
+            days_of_week,
+            interval_days,
+        )
 
     elif action == "take":
         med_id = parsed.get("medication_id")
         name = parsed.get("name")
-        
-        med_id, name, found = await _find_medication_by_name(message, token, med_id, name)
+
+        med_id, name, found = await _find_medication_by_name(
+            message, token, med_id, name
+        )
         if not found:
             return
-        
+
         if not med_id:
             await message.channel.send("Which medication did you take?")
             return
-            
-        resp, status = api_request("post", f"/api/medications/{med_id}/take", token, {})
+
+        # Try to get scheduled time from recent reminder context
+        scheduled_time = await _get_scheduled_time_from_context(message, name)
+
+        # Build request body with scheduled_time if found
+        request_body = {}
+        if scheduled_time:
+            request_body["scheduled_time"] = scheduled_time
+
+        resp, status = api_request(
+            "post", f"/api/medications/{med_id}/take", token, request_body
+        )
         if status == 201:
-            await message.channel.send(f"Logged **{name}**! Great job staying on track.")
+            if scheduled_time:
+                await message.channel.send(
+                    f"✅ Logged **{name}** for {scheduled_time}! Great job staying on track."
+                )
+            else:
+                await message.channel.send(
+                    f"Logged **{name}**! Great job staying on track."
+                )
         else:
             await message.channel.send(f"Error: {resp.get('error', 'Failed to log')}")
 
@@ -92,16 +164,28 @@ async def handle_medication(message, session, parsed):
         med_id = parsed.get("medication_id")
         name = parsed.get("name")
         reason = parsed.get("reason", "Skipped by user")
-        
-        med_id, name, found = await _find_medication_by_name(message, token, med_id, name)
+
+        med_id, name, found = await _find_medication_by_name(
+            message, token, med_id, name
+        )
         if not found:
             return
-        
+
         if not med_id:
             await message.channel.send("Which medication are you skipping?")
             return
-            
-        resp, status = api_request("post", f"/api/medications/{med_id}/skip", token, {"reason": reason})
+
+        # Try to get scheduled time from recent reminder context
+        scheduled_time = await _get_scheduled_time_from_context(message, name)
+
+        # Build request body with scheduled_time if found
+        request_body = {"reason": reason}
+        if scheduled_time:
+            request_body["scheduled_time"] = scheduled_time
+
+        resp, status = api_request(
+            "post", f"/api/medications/{med_id}/skip", token, request_body
+        )
         if status == 201:
             await message.channel.send(f"Skipped **{name}**. {reason}")
         else:
@@ -121,13 +205,15 @@ async def handle_medication(message, session, parsed):
                     taken = item.get("taken_times", [])
                     skipped = item.get("skipped_times", [])
                     is_prn = item.get("is_prn", False)
-                    
+
                     med_name = med.get("name", "Unknown")
                     dosage = f"{med.get('dosage', '')} {med.get('unit', '')}".strip()
-                    
+
                     if is_prn:
                         status_icon = "💊"
-                        lines.append(f"{status_icon} **{med_name}** {dosage} (as needed)")
+                        lines.append(
+                            f"{status_icon} **{med_name}** {dosage} (as needed)"
+                        )
                     elif times:
                         for time in times:
                             if time in taken:
@@ -136,11 +222,15 @@ async def handle_medication(message, session, parsed):
                                 status_icon = "⏭️"
                             else:
                                 status_icon = "⏰"
-                            lines.append(f"{status_icon} **{med_name}** {dosage} at {time}")
+                            lines.append(
+                                f"{status_icon} **{med_name}** {dosage} at {time}"
+                            )
                     else:
                         lines.append(f"💊 **{med_name}** {dosage}")
-                
-                await message.channel.send("**Today's Medications:**\n" + "\n".join(lines))
+
+                await message.channel.send(
+                    "**Today's Medications:**\n" + "\n".join(lines)
+                )
         else:
             error_msg = "Failed to fetch today's schedule"
             await message.channel.send(f"Error: {resp.get('error', error_msg)}")
@@ -157,62 +247,78 @@ async def handle_medication(message, session, parsed):
                     name = med.get("name", "Unknown")
                     qty = med.get("quantity_remaining")
                     refill_date = med.get("refill_date")
-                    
+
                     if qty is not None and qty <= 7:
                         lines.append(f"⚠️ **{name}**: Only {qty} remaining")
                     elif refill_date:
                         lines.append(f"📅 **{name}**: Refill due by {refill_date}")
-                
+
                 await message.channel.send("**Refills Due:**\n" + "\n".join(lines))
         else:
-            await message.channel.send(f"Error: {resp.get('error', 'Failed to fetch refills')}")
+            await message.channel.send(
+                f"Error: {resp.get('error', 'Failed to fetch refills')}"
+            )
 
     elif action == "snooze":
         med_id = parsed.get("medication_id")
         name = parsed.get("name")
         minutes = parsed.get("minutes", 15)
-        
-        med_id, name, found = await _find_medication_by_name(message, token, med_id, name)
+
+        med_id, name, found = await _find_medication_by_name(
+            message, token, med_id, name
+        )
         if not found:
             return
-        
+
         if not med_id:
             await message.channel.send("Which medication reminder should I snooze?")
             return
-            
-        resp, status = api_request("post", f"/api/medications/{med_id}/snooze", token, {"minutes": minutes})
+
+        resp, status = api_request(
+            "post", f"/api/medications/{med_id}/snooze", token, {"minutes": minutes}
+        )
         if status == 200:
             await message.channel.send(f"⏰ Snoozed **{name}** for {minutes} minutes.")
         else:
-            await message.channel.send(f"Error: {resp.get('error', 'Failed to snooze')}")
+            await message.channel.send(
+                f"Error: {resp.get('error', 'Failed to snooze')}"
+            )
 
     elif action == "adherence":
         med_id = parsed.get("medication_id")
         name = parsed.get("name")
-        
+
         if name and not med_id:
             # Look up by name first
-            med_id, name, found = await _find_medication_by_name(message, token, None, name)
+            med_id, name, found = await _find_medication_by_name(
+                message, token, None, name
+            )
             if not found:
                 return
-        
+
         if med_id:
-            resp, status = api_request("get", f"/api/medications/{med_id}/adherence", token)
+            resp, status = api_request(
+                "get", f"/api/medications/{med_id}/adherence", token
+            )
         else:
             resp, status = api_request("get", "/api/medications/adherence", token)
-            
+
         if status == 200:
             if isinstance(resp, list):
                 lines = []
                 for m in resp:
-                    adherence = m.get('adherence_percent')
+                    adherence = m.get("adherence_percent")
                     if adherence is not None:
-                        lines.append(f"- **{m['name']}**: {adherence}% adherence ({m.get('taken', 0)}/{m.get('expected', 0)} doses)")
+                        lines.append(
+                            f"- **{m['name']}**: {adherence}% adherence ({m.get('taken', 0)}/{m.get('expected', 0)} doses)"
+                        )
                     else:
-                        lines.append(f"- **{m['name']}**: PRN medication (no adherence tracking)")
+                        lines.append(
+                            f"- **{m['name']}**: PRN medication (no adherence tracking)"
+                        )
                 await message.channel.send("**Adherence Stats:**\n" + "\n".join(lines))
             else:
-                adherence = resp.get('adherence_percent')
+                adherence = resp.get("adherence_percent")
                 if adherence is not None:
                     await message.channel.send(
                         f"**{resp.get('name')}**:\n"
@@ -221,79 +327,95 @@ async def handle_medication(message, session, parsed):
                         f"- Skipped: {resp.get('skipped', 0)}"
                     )
                 else:
-                    await message.channel.send(f"**{resp.get('name')}**: PRN medication (no adherence tracking)")
+                    await message.channel.send(
+                        f"**{resp.get('name')}**: PRN medication (no adherence tracking)"
+                    )
         else:
-            await message.channel.send(f"Error: {resp.get('error', 'Failed to fetch adherence')}")
+            await message.channel.send(
+                f"Error: {resp.get('error', 'Failed to fetch adherence')}"
+            )
 
     elif action == "delete":
         med_id = parsed.get("medication_id")
         name = parsed.get("name")
         needs_confirmation = parsed.get("needs_confirmation", True)
-        
-        med_id, name, found = await _find_medication_by_name(message, token, med_id, name)
+
+        med_id, name, found = await _find_medication_by_name(
+            message, token, med_id, name
+        )
         if not found:
             return
-        
+
         if not med_id:
             await message.channel.send("Which medication should I delete?")
             return
-        
+
         # Handle confirmation
         if needs_confirmation:
             if "pending_confirmations" not in session:
                 session["pending_confirmations"] = {}
-            
+
             confirmation_id = f"med_delete_{name}"
             session["pending_confirmations"][confirmation_id] = {
                 "action": "delete",
                 "interaction_type": "medication",
                 "medication_id": med_id,
                 "name": name,
-                "needs_confirmation": False  # Skip confirmation next time
+                "needs_confirmation": False,  # Skip confirmation next time
             }
-            
+
             await message.channel.send(
                 f"⚠️ Are you sure you want to delete **{name}**?\n\n"
                 f"This will also delete all logs for this medication.\n\n"
                 f"Reply **yes** to confirm deletion, or **no** to cancel."
             )
             return
-        
+
         # Actually delete
         resp, status = api_request("delete", f"/api/medications/{med_id}", token)
         if status == 200:
             await message.channel.send(f"🗑️ Deleted **{name}** and all its logs.")
         else:
-            await message.channel.send(f"Error: {resp.get('error', 'Failed to delete medication')}")
+            await message.channel.send(
+                f"Error: {resp.get('error', 'Failed to delete medication')}"
+            )
 
     else:
-        await message.channel.send(f"Unknown action: {action}. Try: list, add, delete, take, skip, today, refills, snooze, or adherence.")
+        await message.channel.send(
+            f"Unknown action: {action}. Try: list, add, delete, take, skip, today, refills, snooze, or adherence."
+        )
 
 
-async def _add_medication(message, token, name, dosage, unit, frequency, times, days_of_week, interval_days):
+async def _add_medication(
+    message, token, name, dosage, unit, frequency, times, days_of_week, interval_days
+):
     """Helper to add a medication."""
     data = {
-        "name": name, 
-        "dosage": dosage, 
-        "unit": unit, 
-        "frequency": frequency, 
-        "times": times
+        "name": name,
+        "dosage": dosage,
+        "unit": unit,
+        "frequency": frequency,
+        "times": times,
     }
-    
+
     # Add optional fields if present
     if days_of_week:
         data["days_of_week"] = days_of_week
     if interval_days:
         data["interval_days"] = interval_days
-        
+
     resp, status = api_request("post", "/api/medications", token, data)
     if status == 201:
         schedule_desc = _format_schedule(frequency, times, days_of_week, interval_days)
-        await message.channel.send(f"✅ Added **{name}** ({dosage} {unit}) - {schedule_desc}")
+        await message.channel.send(
+            f"✅ Added **{name}** ({dosage} {unit}) - {schedule_desc}"
+        )
     else:
-        error_msg = resp.get('error', 'Failed to add medication')
+        error_msg = resp.get("error", "Failed to add medication")
         if "invalid input syntax" in error_msg.lower():
-            await message.channel.send(f"Error: I couldn't understand the schedule format. Try saying it differently, like 'every day at 9am' or 'on Monday and Wednesday at 8pm'.")
+            await message.channel.send(
+                f"Error: I couldn't understand the schedule format. Try saying it differently, like 'every day at 9am' or 'on Monday and Wednesday at 8pm'."
+            )
         else:
             await message.channel.send(f"Error: {error_msg}")
 
@@ -302,41 +424,47 @@ async def _find_medication_by_name(message, token, med_id, name):
     """Helper to find medication by name if ID not provided. Returns (med_id, name, found)."""
     if med_id:
         return med_id, name, True
-    
+
     if not name:
         return None, None, True  # Will prompt user later
-    
+
     # Look up medication by name
     resp, status = api_request("get", "/api/medications", token)
     if status != 200:
         await message.channel.send("Error looking up medication. Please try again.")
         return None, None, False
-    
+
     meds = resp if isinstance(resp, list) else []
-    
+
     # Try exact match first
-    matching = [m for m in meds if m['name'].lower() == name.lower()]
-    
+    matching = [m for m in meds if m["name"].lower() == name.lower()]
+
     # Then try partial match
     if not matching:
-        matching = [m for m in meds if name.lower() in m['name'].lower() or m['name'].lower() in name.lower()]
-    
+        matching = [
+            m
+            for m in meds
+            if name.lower() in m["name"].lower() or m["name"].lower() in name.lower()
+        ]
+
     if not matching:
-        await message.channel.send(f"❌ I couldn't find a medication called '{name}'.\n\nYour medications:\n" + 
-            "\n".join([f"- {m['name']}" for m in meds[:10]]))
-        return None, None, False
-    
-    if len(matching) > 1:
-        # Multiple matches - show list
-        lines = [f"{i+1}. {m['name']}" for i, m in enumerate(matching[:5])]
         await message.channel.send(
-            f"🔍 I found multiple medications matching '{name}':\n" + 
-            "\n".join(lines) + 
-            "\n\nPlease specify which one (e.g., 'take 1' or say the full name)."
+            f"❌ I couldn't find a medication called '{name}'.\n\nYour medications:\n"
+            + "\n".join([f"- {m['name']}" for m in meds[:10]])
         )
         return None, None, False
-    
-    return matching[0]['id'], matching[0]['name'], True
+
+    if len(matching) > 1:
+        # Multiple matches - show list
+        lines = [f"{i + 1}. {m['name']}" for i, m in enumerate(matching[:5])]
+        await message.channel.send(
+            f"🔍 I found multiple medications matching '{name}':\n"
+            + "\n".join(lines)
+            + "\n\nPlease specify which one (e.g., 'take 1' or say the full name)."
+        )
+        return None, None, False
+
+    return matching[0]["id"], matching[0]["name"], True
 
 
 def _format_schedule(frequency, times, days_of_week, interval_days):
@@ -346,7 +474,7 @@ def _format_schedule(frequency, times, days_of_week, interval_days):
     elif frequency == "twice_daily":
         return f"twice daily at {', '.join(times)}"
     elif frequency == "specific_days":
-        days_str = ', '.join(days_of_week) if days_of_week else 'certain days'
+        days_str = ", ".join(days_of_week) if days_of_week else "certain days"
         return f"on {days_str} at {', '.join(times)}"
     elif frequency == "every_n_days":
         return f"every {interval_days} days at {', '.join(times)}"
@@ -359,6 +487,7 @@ def _format_schedule(frequency, times, days_of_week, interval_days):
 def api_request(method, endpoint, token, data=None):
     import requests
     import os
+
     API_URL = os.getenv("API_URL", "http://app:5000")
     url = f"{API_URL}{endpoint}"
     headers = {"Content-Type": "application/json", "Authorization": f"Bearer {token}"}
diff --git a/questions.layer4.crossref_answers.json b/questions.layer4.crossref_answers.json
new file mode 100644
index 0000000..f7d8530
--- /dev/null
+++ b/questions.layer4.crossref_answers.json
@@ -0,0 +1,1615 @@
+{
+  "metadata": {
+    "query_file": "questions.layer4.crossref.json",
+    "timestamp": "2026-02-14T00:47:46.873265",
+    "embedding_model": "sentence-transformers/all-minilm-l12-v2",
+    "chat_model": "deepseek/deepseek-v3.2",
+    "total_queries": 264,
+    "processing_time_seconds": 59.331675,
+    "queries_per_second": 4.44956256502113
+  },
+  "books": {
+    "atomichabits": [
+      {
+        "query": "What makes variable rewards effective and when do they become manipulative?",
+        "answer": "Based on the provided context, variable rewards are effective because they increase behavior rather than decrease it when rewards are given intermittently, as discovered by B.F. Skinner in his experiment with rats. The context does not provide information on when variable rewards become manipulative.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:46:50.806292"
+      },
+      {
+        "query": "What makes external rewards effective for ADHD brains specifically?",
+        "answer": "I don't have any information in the provided context about what makes external rewards effective for ADHD brains specifically. The context discusses rewards in general terms related to habit formation, craving, and dopamine, but does not mention ADHD.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:46:50.842311"
+      },
+      {
+        "query": "What time intervals work for ADHD users feeling in sync with clocks?",
+        "answer": "Based on the provided context, there is no information about specific time intervals that work for ADHD users feeling in sync with clocks.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:46:51.917226"
+      },
+      {
+        "query": "Are there examples of when variable rewards became manipulative in practice?",
+        "answer": "Based on the provided context, I do not have information about examples of variable rewards becoming manipulative in practice. The context mentions variable rewards increasing behavior in an experiment with rats, but does not discuss manipulative applications.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:46:56.251727"
+      },
+      {
+        "query": "How does mood state specifically alter decision-making speed and quality?",
+        "answer": "Based strictly on the provided context, I do not have information on how mood state specifically alters decision-making speed and quality. The context discusses that internal states like moods and emotions matter and influence behavior, but it does not provide specific details about their impact on the speed or quality of decisions.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:46:58.871511"
+      },
+      {
+        "query": "What is the relationship between attention span and task complexity?",
+        "answer": "Based on the provided context, there is no information about the relationship between attention span and task complexity.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:01.616738"
+      },
+      {
+        "query": "How do you measure whether users are motivated enough to act on feedback?",
+        "answer": "I don't know. The provided context does not contain information about measuring user motivation to act on feedback.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:04.941412"
+      },
+      {
+        "query": "How do you design for task switching when executive function is impaired?",
+        "answer": "I don't have specific information in the provided context about designing for task switching when executive function is impaired. The context discusses habit formation, mental shortcuts, and cognitive scripts, but doesn't address task switching design for impaired executive function.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:04.957654"
+      },
+      {
+        "query": "Does Barkley provide specific visual time representations that worked for ADHD users?",
+        "answer": "Based on the provided context, I do not have information about Barkley providing specific visual time representations for ADHD users.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:05.104393"
+      },
+      {
+        "query": "How do you identify when to simplify the behavior versus redesign the environment?",
+        "answer": "Based on the provided context, you simplify the behavior when you need to reduce friction and make the action easier, akin to removing a bend in a hose to let water flow naturally. You redesign the environment to make cues more obvious or to make actions easier by fitting habits into the flow of your life, such as placing a gym on your commute path. Environment design is specifically highlighted as an effective way to reduce friction associated with your habits.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:05.865303"
+      },
+      {
+        "query": "How do you create compelling reasons to return without using dark patterns?",
+        "answer": "Based on the provided context, you can create compelling reasons to return by making the habit attractive through positive associations and strategic design:\n\n1. **Make the habit attractive** by associating it with positive feelings. The context states: \"Habits are attractive when we associate them with positive feelings.\"\n2. **Use temptation bundling**: \"Pair an action you want to do with an action you need to do.\"\n3. **Create a motivation ritual**: \"by doing something you enjoy immediately before a difficult habit.\"\n4. **Join a culture** where the desired behavior is normal and valued.\n\nThese methods focus on creating genuine appeal without manipulation.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:07.111194"
+      },
+      {
+        "query": "How do you measure if feedback is actionable versus just informative?",
+        "answer": "Based on the provided context, there is no information about measuring feedback as actionable versus informative. Therefore, I don't know.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:07.135921"
+      },
+      {
+        "query": "Are there case studies of friction levels that broke bad habits without causing dropout?",
+        "answer": "Based solely on the provided context, there is no specific mention of case studies examining friction levels that broke bad habits without causing dropout. The context discusses reducing friction to make good habits easier and increasing friction to make bad habits harder, but it does not detail case studies on this specific balance.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:10.711783"
+      },
+      {
+        "query": "What is the minimum viable action for common behaviors like exercise or sleep routines?",
+        "answer": "Based on the provided context, the minimum viable action for common behaviors is to **downscale your habits until they can be done in two minutes or less**, as stated in the Two-Minute Rule (3.4).",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:11.325799"
+      },
+      {
+        "query": "Are there examples of progress bars that motivated versus demotivated users?",
+        "answer": "I don't know.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:13.172633"
+      },
+      {
+        "query": "What is the exact capacity of working memory in items and duration?",
+        "answer": "Based on the provided context, there is no information about the exact capacity of working memory in terms of items or duration.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:14.091665"
+      },
+      {
+        "query": "What percentage of decisions are made unconsciously versus consciously?",
+        "answer": "I don't know.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:14.468791"
+      },
+      {
+        "query": "Does the book provide specific time intervals for ADHD users syncing with clocks?",
+        "answer": "Based strictly on the provided context, there is no information about specific time intervals for ADHD users syncing with clocks. The context discusses habit formation, implementation intentions, and time management strategies, but does not mention ADHD or clock-syncing intervals. Therefore, I don't know.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:15.814475"
+      },
+      {
+        "query": "What is the working memory capacity for ADHD versus neurotypical users?",
+        "answer": "I don't know. The provided context from \"atomichabits.embeddings\" does not contain information about the working memory capacity for ADHD versus neurotypical users.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:17.632119"
+      },
+      {
+        "query": "Does the book specify working memory capacity differences between ADHD and neurotypical?",
+        "answer": "I don't know.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:18.198698"
+      },
+      {
+        "query": "Does the book provide examples of minimum viable actions that worked for exercise or sleep?",
+        "answer": "Based on the provided context, the book mentions the Two-Minute Rule as a strategy for making habits easy, which is related to the concept of minimum viable actions. Specifically, it discusses making the cues of bad habits difficult and using one-time actions to lead to better habits. For sleep, the context describes a strategy where individuals were told to only go to bed when tired and to sit in a different room if they couldn't fall asleep, thereby associating the bed strictly with sleeping. This is an example of a minimum viable action for improving sleep habits. However, the context does not provide a specific example of a minimum viable action for exercise.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:18.252030"
+      },
+      {
+        "query": "How does emotion dysregulation affect interaction with error messages?",
+        "answer": "Based on the provided context, I do not have information on how emotion dysregulation specifically affects interaction with error messages. The context discusses the role of emotions and cravings in prompting action and changing internal states, but it does not address error messages or the concept of emotion dysregulation in that context.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:19.029384"
+      },
+      {
+        "query": "How do you design for recognition without accidentally creating false memories?",
+        "answer": "Based on the provided context, I don't have information on how to design for recognition without accidentally creating false memories. The context discusses how the brain learns from repeated experiences, encodes information automatically, and how habits become automatic through practice, but it does not address the design of recognition systems or the prevention of false memories.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:21.809793"
+      },
+      {
+        "query": "What makes progress bars motivating versus anxiety-inducing?",
+        "answer": "I don't know. The provided context does not contain information about progress bars being motivating or anxiety-inducing.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:22.549542"
+      },
+      {
+        "query": "How do you measure if something is providing intrinsic versus extrinsic motivation?",
+        "answer": "Based on the provided context, I don't have information on how to measure whether something is providing intrinsic versus extrinsic motivation. The context discusses the concepts and effects of intrinsic motivation (like identity and pride) and extrinsic rewards, but does not provide methods for measuring them.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:25.536153"
+      },
+      {
+        "query": "What specific friction increases work for breaking bad habits without causing abandonment?",
+        "answer": "Based on the provided context, specific friction increases that work for breaking bad habits include:\n\n1. **Increasing the number of steps between you and your bad habits** (Inversion of the 3rd Law: Make It Difficult, point 3.6).\n2. **Using a commitment device** to restrict your future choices to the ones that benefit you (Inversion of the 3rd Law, point 3.7).\n3. **Making the cues of bad habits difficult** (as mentioned in the summary of the 3rd Law of Behavior Change).\n\nThe context provides examples such as hiding beer in the back of the fridge to drink less and deleting social media apps from your phone to reduce usage, noting that \"a little bit of friction can be the difference between sticking with a good habit or sliding into a bad one.\" These tricks are described as unlikely to curb a true addiction but effective for many people in preventing unwanted behavior by adding small friction.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:27.488908"
+      },
+      {
+        "query": "Does the book provide any examples or case studies of notification habituation timeframes?",
+        "answer": "Based strictly on the provided context, I do not have information about whether the book provides examples or case studies of notification habituation timeframes.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:28.598422"
+      },
+      {
+        "query": "How do you externalize executive functions through interface design?",
+        "answer": "Based on the provided context, there is no information about externalizing executive functions through interface design. Therefore, I don't know.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:29.714262"
+      },
+      {
+        "query": "What percentage range is given for unconscious versus conscious decisions?",
+        "answer": "Based on the provided context, there is no specific percentage range given for unconscious versus conscious decisions. The context discusses nonconscious processes and habits but does not quantify them with percentages.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:29.902839"
+      },
+      {
+        "query": "Are there examples of first-time experiences that led to high retention?",
+        "answer": "Based on the provided context, there is no information about examples of first-time experiences leading to high retention.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:32.120204"
+      },
+      {
+        "query": "How far in advance should reminders trigger before deadlines for ADHD users?",
+        "answer": "I don't know. The provided context does not contain information about how far in advance reminders should trigger before deadlines for ADHD users.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:34.489791"
+      },
+      {
+        "query": "Does the book specify optimal animation duration or speed for attention?",
+        "answer": "Based strictly on the provided context from \"atomichabits.embeddings,\" there is no information about optimal animation duration or speed for attention. I don't know.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:37.438442"
+      },
+      {
+        "query": "How long does it take for users to habituate to repetitive notifications?",
+        "answer": "Based on the provided context, there is no information about users habituating to repetitive notifications. The context discusses habit formation in general, focusing on repetition frequency rather than time, but does not address notifications specifically. Therefore, I don't know.",
+        "book": "atomichabits.embeddings",
+        "timestamp": "2026-02-14T00:47:40.702315"
+      }
+    ],
+    "100things": [
+      {
+        "query": "What is the working memory capacity for ADHD versus neurotypical users?",
+        "answer": "I don't know. The provided context does not contain information comparing the working memory capacity of users with ADHD versus neurotypical users.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:51.480312"
+      },
+      {
+        "query": "What is the maximum acceptable feedback delay before users perceive unresponsiveness?",
+        "answer": "Based on the provided context, there is no specific information about the maximum acceptable feedback delay before users perceive unresponsiveness.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:52.556986"
+      },
+      {
+        "query": "Does the book describe techniques to avoid the decoy effect while presenting options?",
+        "answer": "I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:52.866290"
+      },
+      {
+        "query": "What specific haptic patterns work for confirmation versus alerts?",
+        "answer": "Based on the provided context, there is no information about specific haptic patterns for confirmation versus alerts. The context discusses visual cues, affordances, and unconscious decision-making, but does not cover haptic feedback patterns. Therefore, I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:53.596239"
+      },
+      {
+        "query": "How do you identify when to simplify the behavior versus redesign the environment?",
+        "answer": "Based on the provided context, I don't have information on how to identify when to simplify behavior versus redesign the environment. The context discusses affordances, error handling, and user reactions, but does not provide criteria for making this specific design decision.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:53.605428"
+      },
+      {
+        "query": "Are there examples of habit stacking with inconsistent anchors?",
+        "answer": "Based on the provided context, there is no information about examples of habit stacking with inconsistent anchors.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:55.000763"
+      },
+      {
+        "query": "Does Krug provide specific numbers for choice thresholds that trigger paralysis?",
+        "answer": "Based solely on the provided context, there is no mention of Krug or specific numbers for choice thresholds that trigger paralysis. The context discusses research by Sheena Iyengar and others on choice overload, but it does not reference anyone named Krug providing specific numerical thresholds.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:55.066724"
+      },
+      {
+        "query": "How do you design for one-handed mobile use specifically?",
+        "answer": "Based on the provided context, I don't have specific information on how to design for one-handed mobile use. The context discusses topics like inattention blindness, how vision works, and attention problems while driving, but it does not address design considerations for one-handed mobile device interaction.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:56.512684"
+      },
+      {
+        "query": "How do defaults interact with making habits obvious?",
+        "answer": "Based on the provided context, there is no information about how defaults interact with making habits obvious. The context discusses defaults and habit formation as separate topics without connecting them.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:57.620060"
+      },
+      {
+        "query": "What makes a cue noticeable enough to trigger the habit loop in a digital interface?",
+        "answer": "Based on the provided context, I don't have information on what makes a cue noticeable enough to trigger a habit loop in a digital interface. The context discusses habit formation in general and mentions cues in relation to dopamine loops and information-seeking, but does not specify the characteristics of effective cues within digital interfaces.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:57.715438"
+      },
+      {
+        "query": "How do different emotions specifically alter decision-making patterns?",
+        "answer": "Based on the provided context, different emotions like happiness and sadness specifically alter decision-making patterns by influencing whether an intuitive or deliberative strategy leads to higher product valuation.\n\nThe research by Marieke de Vries (2008) shows:\n*   People in a **happy mood** will rate a product as more valuable if they make the decision **quickly based on their first feelings** (intuitive condition).\n*   People in a **sad mood** will rate a product as more valuable if they make the decision in a **more deliberate way**, evaluating pros and cons (deliberative condition).\n\nTherefore, the context indicates that mood (emotion) influences which decision-making strategy (intuitive vs. deliberative) results in a higher perceived value for a product or service.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:57.976360"
+      },
+      {
+        "query": "How do you externalize executive functions through interface design?",
+        "answer": "I don't have enough information in the provided context to answer that question. The context discusses various topics like working memory, brain activity, autonomy, motivation, and design principles, but it doesn't specifically address how to externalize executive functions through interface design.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:58.822786"
+      },
+      {
+        "query": "How do you design for task switching when executive function is impaired?",
+        "answer": "Based on the provided context, I don't have information on how to design for task switching when executive function is impaired. The context discusses working memory, mind wandering, flow states, and the effects of stress and multitasking on performance, but it does not address design considerations for impaired executive function.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:59.685705"
+      },
+      {
+        "query": "How do you measure if a habit is easy enough according to the third law?",
+        "answer": "Based on the provided context, there is no information on how to measure if a habit is \"easy enough\" according to a \"third law.\" The context discusses the time it takes to form habits and the impact of complexity and consistency, but does not mention any specific laws or a measurement for ease. Therefore, I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:46:59.760615"
+      },
+      {
+        "query": "Does Barkley provide specific visual time representations that worked for ADHD users?",
+        "answer": "I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:00.417447"
+      },
+      {
+        "query": "Are there examples of preserving social norms with optional payment systems?",
+        "answer": "I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:01.230843"
+      },
+      {
+        "query": "Are there examples of first-time experiences that led to high retention?",
+        "answer": "Based on the provided context, there is no specific mention of examples of first-time experiences leading to high retention. The context discusses memory in general, including concepts like the Forgetting Curve, flashbulb memories, and reconstructive memory, but it does not provide concrete examples of first-time experiences and their retention rates.\n\nTherefore, I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:01.466123"
+      },
+      {
+        "query": "How do you design habit stacking when users lack reliable daily anchor habits?",
+        "answer": "Based on the provided context, there is no information about designing habit stacking or addressing situations where users lack reliable daily anchor habits. Therefore, I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:03.269644"
+      },
+      {
+        "query": "Does the book compare digital versus physical reward effectiveness?",
+        "answer": "Based on the provided context, the book does compare digital versus physical reward effectiveness, specifically regarding monetary rewards. It states that giving people monetary rewards can backfire because they may come to rely on them and then be unwilling to work unless there is a monetary incentive. This is presented in contrast to other types of rewards or work motivations.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:03.638558"
+      },
+      {
+        "query": "How far in advance should reminders trigger before deadlines for ADHD users?",
+        "answer": "I don't know. The provided context does not contain information about how far in advance reminders should trigger before deadlines, nor does it specifically address strategies for ADHD users.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:04.957462"
+      },
+      {
+        "query": "What is good enough and how do you communicate that standard?",
+        "answer": "Based on the provided context, there is no information about what constitutes \"good enough\" or how to communicate that standard.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:04.979931"
+      },
+      {
+        "query": "Are there examples of haptic patterns used for confirmation versus alerts?",
+        "answer": "I don't have information about haptic patterns for confirmation versus alerts in the provided context.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:05.574759"
+      },
+      {
+        "query": "What time intervals work for ADHD users feeling in sync with clocks?",
+        "answer": "I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:07.608285"
+      },
+      {
+        "query": "How many choices on a page trigger decision paralysis?",
+        "answer": "Based on the provided context, decision paralysis occurs when there are too many choices, specifically when the number of options exceeds what people can process effectively. The context indicates that people can decide from among only three or four things at a time. The \"jam\" study example shows a significant drop in purchases (from 31% to 3%) when choices increased from six to twenty-four, suggesting that even six choices may be too many. The key takeaway advises limiting choices to three or four. Therefore, having more than three or four choices on a page is likely to trigger decision paralysis.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:07.657267"
+      },
+      {
+        "query": "How do you measure if feedback is actionable versus just informative?",
+        "answer": "Based on the provided context, I don't have information on how to measure if feedback is actionable versus just informative. The context discusses unconscious decision-making processes, social validation, and how people value products, but does not address the evaluation of feedback in this way.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:08.043179"
+      },
+      {
+        "query": "Does the book provide specific time intervals for ADHD users syncing with clocks?",
+        "answer": "I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:09.462694"
+      },
+      {
+        "query": "Are there examples showing what makes navigation feel automatic versus requiring thought?",
+        "answer": "Based solely on the provided context, there is no information about what makes navigation feel automatic versus requiring thought.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:10.178336"
+      },
+      {
+        "query": "What makes the perfect toggle - size, feedback, animation timing?",
+        "answer": "Based on the provided context, there is no information about what makes the perfect toggle regarding size, feedback, or animation timing. I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:10.665215"
+      },
+      {
+        "query": "How do you avoid the decoy effect while still presenting options?",
+        "answer": "Based on the provided context, I don't have information about the decoy effect or how to avoid it while presenting options. The context discusses limiting choices, progressive disclosure, and how people process choices, but does not mention the decoy effect.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:11.155962"
+      },
+      {
+        "query": "Does Schwartz provide a specific optimal number of choices for any decision types?",
+        "answer": "Based strictly on the provided context, there is no mention of Schwartz or any specific optimal number of choices provided by Schwartz for decision types. The context discusses research by Sheena Iyengar and takeaways about limiting choices, but does not reference Schwartz in relation to an optimal number.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:11.423434"
+      },
+      {
+        "query": "How do users decide what is the first reasonable option when satisficing?",
+        "answer": "Based on the provided context, users decide what is the first reasonable option when satisficing by scanning and clicking on the first link that catches their interest or vaguely resembles what they are looking for, rather than reading everything or analyzing all options.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:11.467079"
+      },
+      {
+        "query": "How do you frame streaks as identity markers rather than performance metrics?",
+        "answer": "Based on the provided context, I don't have information on how to frame streaks as identity markers rather than performance metrics. The context discusses habit formation, attention, rewards, and memory, but does not address the specific concept of reframing streaks.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:12.480587"
+      },
+      {
+        "query": "How do you automate habits without removing the sense of agency?",
+        "answer": "Based on the provided context, I don't have information on how to automate habits while preserving a sense of agency. The context discusses habit formation, autonomy, and motivation, but does not address the specific method of automating habits without removing agency.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:13.171173"
+      },
+      {
+        "query": "What makes external rewards effective for ADHD brains specifically?",
+        "answer": "Based on the provided context, there is no specific information about what makes external rewards effective for ADHD brains. The context discusses rewards, dopamine, and motivation in general, but does not mention ADHD. Therefore, I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:13.171298"
+      },
+      {
+        "query": "How long do expectation effects persist after initial priming?",
+        "answer": "I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:13.358026"
+      },
+      {
+        "query": "How do reversible choices affect satisfaction compared to final ones?",
+        "answer": "I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:15.287913"
+      },
+      {
+        "query": "What specific defaults reduce choice overload most effectively?",
+        "answer": "Based on the provided context, there is no information about specific defaults that reduce choice overload. The context discusses defaults in relation to reducing the amount of work, but does not connect them to the issue of choice overload or specify which defaults are most effective for that purpose.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:15.435515"
+      },
+      {
+        "query": "Does Clear provide examples of UI or design reinforcing identity implicitly?",
+        "answer": "Based on the provided context, there is no information about Clear providing examples of UI or design reinforcing identity implicitly. The context discusses general principles of how people see, read, remember, and think in relation to design, but does not mention the specific application \"Clear\" or examples related to identity reinforcement. Therefore, I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:16.083423"
+      },
+      {
+        "query": "Does Ariely provide examples of expectation priming duration from his experiments?",
+        "answer": "Based on the provided context, there is no mention of Ariely or examples of expectation priming duration from his experiments. Therefore, I do not have that information.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:16.404494"
+      },
+      {
+        "query": "How do you design for satisficing users versus maximizing users?",
+        "answer": "Based on the provided context, I can only give a partial answer. For satisficing users, you should design websites for scanning, not reading, as they will glance at a page, scan text, and click on the first link that catches their interest. The context does not provide information on how to design for maximizing users.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:16.895510"
+      },
+      {
+        "query": "What priming techniques work in digital interfaces versus physical spaces?",
+        "answer": "Based on the provided context, I don't have information about specific priming techniques that work in digital interfaces versus physical spaces. The context discusses concepts like affordances, inattention blindness, and cognitive loads in design, but does not compare priming techniques across these two environments.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:17.427189"
+      },
+      {
+        "query": "Are there case studies of friction levels that broke bad habits without causing dropout?",
+        "answer": "Based on the provided context from **100things.embeddings**, there is **no information** about case studies specifically examining friction levels used to break bad habits without causing dropout. The context focuses on how habits are formed (e.g., time required, consistency, small steps) and related motivational factors, but does not address interventions or case studies for breaking habits through friction.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:18.256819"
+      },
+      {
+        "query": "Does the book provide examples of minimum viable actions that worked for exercise or sleep?",
+        "answer": "Based on the provided context, I do not have information about whether the book provides examples of minimum viable actions that worked for exercise or sleep. The context discusses research on sleep and insight, and synchronous activities, but does not mention \"minimum viable actions\" in relation to exercise or sleep.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:19.465377"
+      },
+      {
+        "query": "Are there specific scenarios where the endowment effect was stronger or weaker?",
+        "answer": "Based on the provided context, there is no information about specific scenarios where the endowment effect was stronger or weaker. The context only discusses that people value a product more highly when it's physically in front of them, but it does not detail conditions that modulate the strength of this effect.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:19.654766"
+      },
+      {
+        "query": "Should you hide metrics during plateaus or make the plateau itself visible?",
+        "answer": "Based on the provided context, I don't have information about whether to hide metrics during plateaus or make the plateau itself visible. The context discusses habit formation curves and mastery as asymptotes, but does not address strategies for presenting metrics or plateaus.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:21.095429"
+      },
+      {
+        "query": "How long do plateaus typically last for different habit types?",
+        "answer": "Based on the provided context, I don't know how long plateaus typically last for different habit types. The context mentions that people hit a plateau in automaticity, forming an asymptote curve, but it does not specify the duration of these plateaus for different types of habits.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:22.255162"
+      },
+      {
+        "query": "How do you create compelling reasons to return without using dark patterns?",
+        "answer": "Based on the provided context, there is no information on creating compelling reasons to return without using dark patterns. The context discusses unconscious decision-making, forming habits, learning from examples, and errors, but does not address strategies for encouraging return visits or the concept of dark patterns. Therefore, I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:22.264312"
+      },
+      {
+        "query": "Does the book specify when instructions helped versus harmed in actual tests?",
+        "answer": "Based on the provided context, the book does not specify when instructions helped versus harmed in actual tests. The context discusses instructions in relation to a study on helping behavior, but it does not detail the conditions under which those instructions were beneficial or detrimental.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:22.269604"
+      },
+      {
+        "query": "How do you reduce anticipated regret through design?",
+        "answer": "Based on the provided context, I don't have information on how to reduce anticipated regret through design. The context mentions research on understanding others' regret and how people value a product more when it's physically present, but it does not discuss design strategies for reducing anticipated regret.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:22.647887"
+      },
+      {
+        "query": "How does emotion dysregulation affect interaction with error messages?",
+        "answer": "Based on the provided context, there is no information about emotion dysregulation or its effect on interaction with error messages.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:22.741226"
+      },
+      {
+        "query": "What mapping patterns work best for timeline controls and why?",
+        "answer": "I don't know. The provided context does not contain information about mapping patterns for timeline controls or why certain patterns might work best.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:22.832026"
+      },
+      {
+        "query": "What is the actual optimal number of choices for different decision types?",
+        "answer": "Based on the provided context, the optimal number of choices is **three or four**. The text states: \"If possible, limit the number of choices to three or four.\" It also explains that people can only decide from among three or four things at a time, paralleling their memory limits. The context does not specify different optimal numbers for different types of decisions.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:23.067547"
+      },
+      {
+        "query": "How do you measure if something makes users think too much?",
+        "answer": "Based on the provided context, there is no information on how to measure if something makes users think too much.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:23.544366"
+      },
+      {
+        "query": "What is the minimum time between response and reward before the feedback loop breaks?",
+        "answer": "I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:24.293523"
+      },
+      {
+        "query": "How do you test for cognitive friction in usability testing?",
+        "answer": "I don't have information in the provided context about how to test for cognitive friction in usability testing. The context discusses cognitive, visual, and motor loads, and trade-offs between them, but does not mention specific testing methods for cognitive friction.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:25.558272"
+      },
+      {
+        "query": "What is the minimum viable action for common behaviors like exercise or sleep routines?",
+        "answer": "Based on the provided context, there is no specific information about the \"minimum viable action\" for common behaviors like exercise or sleep routines. The context discusses habit formation in general, including that complex behaviors like exercise take longer to become automatic, but it does not define or describe a minimum viable action for these specific behaviors.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:25.647300"
+      },
+      {
+        "query": "Do digital rewards like checkmarks and animations satisfy the same as physical rewards?",
+        "answer": "I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:25.731166"
+      },
+      {
+        "query": "Are specific mapping patterns recommended for time-based controls?",
+        "answer": "I don't know.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:28.450186"
+      },
+      {
+        "query": "Does the book specify working memory capacity differences between ADHD and neurotypical?",
+        "answer": "Based on the provided context, there is no information about working memory capacity differences between ADHD and neurotypical individuals.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:28.718297"
+      },
+      {
+        "query": "What makes navigation require thinking versus being automatic?",
+        "answer": "Based on the provided context, I don't have information about what specifically makes navigation require thinking versus being automatic. The context discusses automaticity in the context of well-practiced skills and error-prone sequences, but does not address navigation.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:30.424629"
+      },
+      {
+        "query": "Does the book describe how to measure or test for erroneous mental models?",
+        "answer": "Based on the provided context, there is no description of how to measure or test for erroneous mental models. The context discusses the definition of mental models, their origins, and their role in design, but does not include methods for measuring or testing them.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:30.750159"
+      },
+      {
+        "query": "Does the book specify how many habits in a stack before fragility occurs?",
+        "answer": "No, the provided context does not specify a number of habits in a stack before fragility occurs.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:30.938669"
+      },
+      {
+        "query": "What specific friction increases work for breaking bad habits without causing abandonment?",
+        "answer": "Based on the provided context, there is no information about specific friction that increases work for breaking bad habits without causing abandonment.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:31.977903"
+      },
+      {
+        "query": "What makes temptation bundling effective and when does it fail?",
+        "answer": "I don't know. The provided context does not contain information about temptation bundling, its effectiveness, or when it fails.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:32.110496"
+      },
+      {
+        "query": "Can you design UI that reinforces identity without explicitly stating it?",
+        "answer": "Based on the provided context, I don't have information about designing UI to reinforce identity without explicitly stating it. The context discusses mental models, conceptual models, attention, and information processing, but does not address the topic of identity reinforcement in UI design.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:32.773874"
+      },
+      {
+        "query": "What is the threshold where instructions become necessary versus harmful?",
+        "answer": "Based on the provided context, I don't have information about a specific threshold where instructions become necessary versus harmful. The context discusses performance, stress, errors, and memory, but does not address the point at which instructions shift from being helpful to detrimental.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:34.009441"
+      },
+      {
+        "query": "What specific environmental cues make good habits obvious in digital spaces?",
+        "answer": "Based on the provided context, I don't have information on specific environmental cues that make good habits obvious in digital spaces. The context discusses habit formation in general and mentions digital routines like checking email or Facebook, but it does not detail the specific cues that make these habits obvious.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:37.709917"
+      },
+      {
+        "query": "Are there examples of Poka-Yoke implementation in digital interfaces?",
+        "answer": "I don't know. The provided context does not mention \"Poka-Yoke\" or provide examples of its implementation in digital interfaces.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:39.609612"
+      },
+      {
+        "query": "How do you implement Poka-Yoke principles in digital interfaces?",
+        "answer": "I don't know. The provided context does not contain information about Poka-Yoke principles or their implementation in digital interfaces.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:40.096651"
+      },
+      {
+        "query": "What makes the endowment effect stronger or weaker?",
+        "answer": "Based on the provided context, I don't have information about what makes the endowment effect stronger or weaker. The context does not mention the endowment effect.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:40.608401"
+      },
+      {
+        "query": "What is the threshold where choice becomes too much?",
+        "answer": "Based on the provided context, the threshold where choice becomes too much is when there are more than three or four options. The text advises to limit choices to three or four, and if more options are necessary, to present them in a progressive way (e.g., choosing from a subset after an initial selection).",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:40.976677"
+      },
+      {
+        "query": "What happens when the anchor habit is inconsistent or varies in timing?",
+        "answer": "Based on the provided context, there is no specific information about \"anchor habits\" or what happens when they are inconsistent or vary in timing. The context discusses general habit formation and the effects of missing days, but does not mention anchor habits.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:41.800408"
+      },
+      {
+        "query": "How do you measure whether users are motivated enough to act on feedback?",
+        "answer": "Based on the provided context, I don't have information on specific methods for measuring whether users are motivated enough to act on feedback.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:42.692732"
+      },
+      {
+        "query": "How do people update their mental models after encountering errors?",
+        "answer": "Based on the provided context, people update their mental models primarily through **training** and **experience**, but the context does not explicitly state that encountering errors directly updates mental models.\n\nThe context discusses:\n1. **Training as a method to change mental models**: If a product's conceptual model doesn't match the user's mental model, training (like a short video) can be used to adjust the mental model to fit the design.\n2. **Mental models change with experience**: It states \"mental models are subject to change\" and that they \"start to change and adjust to reflect your experience.\"\n3. **Error correction strategies**: When people encounter errors, they may use strategies like \"systematic exploration\" to correct them (e.g., methodically trying menu options). However, the context does not link this error correction process directly to an update of the underlying mental model.\n\nTherefore, while experience (which could include encountering errors) leads to adjustment, the specific mechanism of updating mental models *after errors* is not detailed in the provided excerpts.",
+        "book": "100things.embeddings",
+        "timestamp": "2026-02-14T00:47:46.799677"
+      }
+    ],
+    "designofeverydaythings": [
+      {
+        "query": "Are there examples of when variable rewards became manipulative in practice?",
+        "answer": "I don't know.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:46:51.521125"
+      },
+      {
+        "query": "What makes variable rewards effective and when do they become manipulative?",
+        "answer": "Based on the provided context, there is no information about variable rewards, their effectiveness, or when they become manipulative. I don't know.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:46:52.819345"
+      },
+      {
+        "query": "Are there examples showing what makes navigation feel automatic versus requiring thought?",
+        "answer": "Yes, based on the provided context, there are examples illustrating what makes navigation feel automatic versus requiring thought.\n\n**Automatic navigation** occurs with well-learned skills where conscious oversight is minimized. The context states: \"Well-learned skills bypass the need for conscious oversight and control... Continual practice automates the action cycle.\" A specific example given is a **skilled driver** making a left turn: \"If I am a skilled driver, I don\u2019t have to give much conscious attention to specify or perform the action sequence. I think 'left' and smoothly execute the required action.\"\n\n**Navigation requiring thought** happens during initial learning or in unexpected situations. The same driving example contrasts this: \"But if I am just learning to drive, I have to think about each separate component of the action.\" The context lists the conscious steps a novice must perform, such as checking for cars, operating pedals, managing turn signals, and dividing visual attention.\n\nThe principle is summarized by a quote from Alfred North Whitehead: \"Civilization advances by extending the number of important operations which we can perform without thinking about them.\"",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:46:58.171707"
+      },
+      {
+        "query": "How do you measure if something makes users think too much?",
+        "answer": "Based on the provided context, I don't have information on specific methods for measuring if something makes users think too much. The context discusses levels of processing (visceral, behavioral, reflective) and the interplay between conscious and subconscious thought, but it does not detail measurement techniques for cognitive load.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:46:58.289575"
+      },
+      {
+        "query": "Are there examples of haptic patterns used for confirmation versus alerts?",
+        "answer": "Based on the provided context, there is no information about haptic patterns being used for confirmation versus alerts. The context discusses auditory and haptic modalities for warning systems in automobiles in a list, but does not provide specific examples or distinctions between their use for confirmation versus alerts.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:46:58.670785"
+      },
+      {
+        "query": "What makes navigation require thinking versus being automatic?",
+        "answer": "Based on the provided context, navigation requires thinking (conscious thought) when it is being learned or when dealing with unexpected situations. Once a skill like driving is well-learned and practiced, it becomes automatic (subconscious), minimizing the need for conscious reasoning. Experts only require conscious attention for especially difficult or unexpected situations.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:46:59.133901"
+      },
+      {
+        "query": "What is the relationship between attention span and task complexity?",
+        "answer": "Based on the provided context, there is no direct information about the relationship between attention span and task complexity.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:00.417313"
+      },
+      {
+        "query": "Does Krug provide specific numbers for choice thresholds that trigger paralysis?",
+        "answer": "I don't know.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:02.831708"
+      },
+      {
+        "query": "How do you design for one-handed mobile use specifically?",
+        "answer": "Based solely on the provided context, there is no specific information on designing for one-handed mobile use. The context discusses general principles like affordances, constraints, mappings, and human-centered design, and mentions touch-sensitive screens lacking physical affordances, but it does not address the specific considerations for one-handed interaction.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:04.958665"
+      },
+      {
+        "query": "How do users decide what is the first reasonable option when satisficing?",
+        "answer": "Based on the provided context, users decide on the first reasonable option when satisficing by using their own conceptual models of the world to determine the perceived causal relationship between the thing being blamed and the result. The word \"perceived\" is critical: the causal relationship does not have to exist; the person simply has to think it is there.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:06.780939"
+      },
+      {
+        "query": "Does the book specify when instructions helped versus harmed in actual tests?",
+        "answer": "I don't know.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:08.097306"
+      },
+      {
+        "query": "How do you design for satisficing users versus maximizing users?",
+        "answer": "Based on the provided context, there is no information about designing for satisficing users versus maximizing users. I don't know.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:08.515062"
+      },
+      {
+        "query": "What makes the perfect toggle - size, feedback, animation timing?",
+        "answer": "Based on the provided context, I do not have information about the specific design characteristics of a perfect toggle, such as its ideal size, feedback mechanisms, or animation timing. The context discusses general design principles like affordances, signifiers, feedback, and mappings, but does not apply these principles to the specific details of toggle switch design.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:09.967064"
+      },
+      {
+        "query": "What is the exact capacity of working memory in items and duration?",
+        "answer": "Based on the provided context, the exact capacity of working memory (short-term memory) is described as ranging from five to seven items traditionally, but from a practical point of view, it is best to think of it as holding only three to five items. Regarding duration, the context does not specify an exact time duration; it only mentions that retention is affected by both time and the number of items, with the number being more important.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:12.360062"
+      },
+      {
+        "query": "Are there examples of progress bars that motivated versus demotivated users?",
+        "answer": "I don't know.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:14.032413"
+      },
+      {
+        "query": "What percentage of decisions are made unconsciously versus consciously?",
+        "answer": "Based solely on the provided context, the exact percentage of decisions made unconsciously versus consciously is not specified. The text discusses how well-learned skills and behaviors become automatic (unconscious) through practice, while conscious thought is required for learning and unexpected situations, but no numerical percentage is given.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:16.071310"
+      },
+      {
+        "query": "How do you test for cognitive friction in usability testing?",
+        "answer": "I don't know.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:16.236765"
+      },
+      {
+        "query": "Does the book specify optimal animation duration or speed for attention?",
+        "answer": "Based on the provided context, there is no mention of optimal animation duration or speed for attention.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:17.195645"
+      },
+      {
+        "query": "Are there examples of Poka-Yoke implementation in digital interfaces?",
+        "answer": "Based solely on the provided context, there are no specific examples of Poka-Yoke implementation in digital interfaces mentioned. The context only describes physical examples of Poka-Yoke, such as using stick-on dots on doors, alignment jigs, asymmetrical screw holes, and covers for emergency switches.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:17.426905"
+      },
+      {
+        "query": "How do you implement Poka-Yoke principles in digital interfaces?",
+        "answer": "Based on the provided context, there is no information about implementing Poka-Yoke principles specifically in digital interfaces. The context discusses Poka-Yoke in physical contexts like manufacturing and home use, applying principles such as signifiers, affordances, mapping, constraints, and forcing functions. However, it does not extend these examples to digital interfaces.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:20.373140"
+      },
+      {
+        "query": "What is the threshold where instructions become necessary versus harmful?",
+        "answer": "Based on the provided context, I do not have information about the threshold where instructions become necessary versus harmful.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:21.155748"
+      },
+      {
+        "query": "How do you design for recognition without accidentally creating false memories?",
+        "answer": "Based on the provided context, I cannot answer how to design for recognition without accidentally creating false memories. The context discusses how memory retrieval is reconstructive and prone to error, including the formation of false memories, but it does not provide design principles or methods to avoid this issue.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:21.208829"
+      },
+      {
+        "query": "How long does it take for users to habituate to repetitive notifications?",
+        "answer": "Based on the provided context, there is no specific information about how long it takes for users to habituate to repetitive notifications.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:21.275009"
+      },
+      {
+        "query": "What specific haptic patterns work for confirmation versus alerts?",
+        "answer": "Based on the provided context, there is no specific information about haptic patterns for confirmation versus alerts. The context discusses auditory feedback and general principles of feedback, but does not detail specific haptic patterns.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:23.106248"
+      },
+      {
+        "query": "How do you measure if something is providing intrinsic versus extrinsic motivation?",
+        "answer": "Based on the provided context, I cannot answer your question about measuring intrinsic versus extrinsic motivation. The context discusses levels of processing (visceral, behavioral, reflective), expectations, and goals (be-goals, do-goals, motor-goals), but it does not address the specific measurement of intrinsic or extrinsic motivation.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:26.505122"
+      },
+      {
+        "query": "Does the book provide any examples or case studies of notification habituation timeframes?",
+        "answer": "No, based on the provided context, there is no mention of examples or case studies specifically about notification habituation timeframes.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:28.837651"
+      },
+      {
+        "query": "How many choices on a page trigger decision paralysis?",
+        "answer": "I don't know.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:30.431156"
+      },
+      {
+        "query": "How does mood state specifically alter decision-making speed and quality?",
+        "answer": "Based on the provided context, there is no specific information on how mood state alters decision-making speed and quality. The context discusses emotional states in relation to processing levels and the action cycle, but does not detail their specific effects on the speed or quality of decisions.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:35.426464"
+      },
+      {
+        "query": "What percentage range is given for unconscious versus conscious decisions?",
+        "answer": "I don't know. The provided context does not give a specific percentage range for unconscious versus conscious decisions.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:36.575146"
+      },
+      {
+        "query": "What makes progress bars motivating versus anxiety-inducing?",
+        "answer": "Based on the provided context, I don't have specific information about progress bars. However, the context discusses general principles about expectations and emotional responses that could relate to your question:\n\n*   **Motivation** arises from a positive affective response when expectations of a positive outcome are met or confirmed by feedback. This can lead to satisfaction, relief, and a feeling of control.\n*   **Anxiety or frustration** arises from negative expectations or when feedback disconfirms expectations, especially if the reasons for failure or possible remedies are unknown.\n\nThe concept of \"flow\" is also mentioned, where an engaging state occurs when the challenge of an activity slightly exceeds one's skill level, requiring full attention. A task that is too difficult relative to skill leads to frustration and anxiety.\n\nWhile these principles about expectations, feedback, and challenge versus skill could apply to the experience of watching a progress bar, the context does not explicitly analyze progress bars themselves.",
+        "book": "designofeverydaythings.embeddings",
+        "timestamp": "2026-02-14T00:47:41.832019"
+      }
+    ],
+    "dontmakemethink": [
+      {
+        "query": "Are there examples of Poka-Yoke implementation in digital interfaces?",
+        "answer": "I don't know.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:46:52.469313"
+      },
+      {
+        "query": "Are specific mapping patterns recommended for time-based controls?",
+        "answer": "Based on the provided context, there is no information about recommended mapping patterns for time-based controls.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:46:54.773742"
+      },
+      {
+        "query": "What mapping patterns work best for timeline controls and why?",
+        "answer": "Based on the provided context, there is no information about mapping patterns for timeline controls. The context discusses design conventions, visual hierarchies, navigation consistency, and the use of Flash, but does not mention timeline controls or their mapping patterns. Therefore, I don't know.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:46:55.669932"
+      },
+      {
+        "query": "What is the actual optimal number of choices for different decision types?",
+        "answer": "Based on the provided context, there is no information about the optimal number of choices for different decision types. The context discusses how people make decisions (like satisficing and muddling through) but does not specify optimal numbers of choices.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:46:58.822233"
+      },
+      {
+        "query": "How do reversible choices affect satisfaction compared to final ones?",
+        "answer": "Based on the provided context, reversible choices (like guessing and using the Back button if wrong) are associated with satisficing and can lead to greater satisfaction because they are efficient, low-risk, and even fun. The context states that with quick page loads, the penalty for a wrong guess is minimal\u2014just a click of the Back button\u2014making satisficing an effective strategy. It also notes that \"Guessing is more fun\" and introduces the pleasant possibility of surprise. In contrast, the context does not provide information about \"final ones\" or directly compare the two.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:47:00.278857"
+      },
+      {
+        "query": "Does the book describe how to measure or test for erroneous mental models?",
+        "answer": "Based on the provided context, I don't know if the book describes how to measure or test for erroneous mental models. The context mentions that people use things without understanding how they work and that they \"muddle through\" with wrong-headed ideas, but it does not specify methods for measuring or testing these mental models.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:47:00.493978"
+      },
+      {
+        "query": "How do you design for one-handed mobile use specifically?",
+        "answer": "I don't know.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:47:01.086337"
+      },
+      {
+        "query": "How do you implement Poka-Yoke principles in digital interfaces?",
+        "answer": "I don't know.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:47:03.449915"
+      },
+      {
+        "query": "How do you reduce anticipated regret through design?",
+        "answer": "Based solely on the provided context, I don't have information on how to reduce anticipated regret through design. The context discusses design practices, decision-making (satisficing), and conflicts between designers and stakeholders, but it does not address the specific concept of \"anticipated regret.\"",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:47:03.569459"
+      },
+      {
+        "query": "Are there examples of haptic patterns used for confirmation versus alerts?",
+        "answer": "Based solely on the provided context, there is no information about haptic patterns being used for confirmation versus alerts. The context discusses visual design elements like noise, accessibility, decision-making, and usability testing, but does not mention haptic patterns.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:47:05.658614"
+      },
+      {
+        "query": "Does Schwartz provide a specific optimal number of choices for any decision types?",
+        "answer": "No, based on the provided context, Schwartz does not provide a specific optimal number of choices for any decision types.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:47:06.780502"
+      },
+      {
+        "query": "What is good enough and how do you communicate that standard?",
+        "answer": "Based on the provided context, I don't have information about what constitutes \"good enough\" or how to communicate that standard.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:47:09.975327"
+      },
+      {
+        "query": "What specific haptic patterns work for confirmation versus alerts?",
+        "answer": "I don't have information about specific haptic patterns for confirmation versus alerts in the provided context. The context discusses visual design elements like background noise, navigation menus, decision-making processes, accessibility solutions, and usability testing, but doesn't mention haptic patterns or tactile feedback systems.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:47:12.356958"
+      },
+      {
+        "query": "What specific defaults reduce choice overload most effectively?",
+        "answer": "Based on the provided context, the specific defaults that reduce choice overload most effectively are not explicitly mentioned. The context discusses strategies like satisficing and the importance of well-integrated design, but it does not list particular default settings or examples that are most effective for reducing choice overload. Therefore, I don't know.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:47:13.171426"
+      },
+      {
+        "query": "How do people update their mental models after encountering errors?",
+        "answer": "Based on the provided context, I don't have information on how people update their mental models after encountering errors. The context discusses how people use things without understanding them and \"muddle through,\" but it does not describe the process of updating mental models following errors.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:47:13.650577"
+      },
+      {
+        "query": "What makes the perfect toggle - size, feedback, animation timing?",
+        "answer": "Based on the provided context, there is no information about what makes the perfect toggle in terms of size, feedback, or animation timing. The context discusses usability principles like \"Don't make me think,\" the importance of testing design decisions, and avoiding distracting elements, but does not specify details about toggle design. Therefore, I don't know.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:47:15.701801"
+      },
+      {
+        "query": "What is the maximum acceptable feedback delay before users perceive unresponsiveness?",
+        "answer": "I don't know. The provided context does not mention specific timing thresholds for feedback delays or user perception of unresponsiveness.",
+        "book": "dontmakemethink.embeddings",
+        "timestamp": "2026-02-14T00:47:19.027697"
+      }
+    ],
+    "designingforbehaviorchange": [
+      {
+        "query": "Does the book specify working memory capacity differences between ADHD and neurotypical?",
+        "answer": "I don't know.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:46:52.578171"
+      },
+      {
+        "query": "Can you design UI that reinforces identity without explicitly stating it?",
+        "answer": "Based on the provided context, the answer is not explicitly stated. The context discusses techniques like narrating to change how users see themselves and associating actions with positive experiences, but it does not specify whether UI can be designed to reinforce identity without explicit statements.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:46:52.861406"
+      },
+      {
+        "query": "How do different emotions specifically alter decision-making patterns?",
+        "answer": "Based on the provided context, emotions are mentioned as part of the intuitive \"gut feeling\" that serves as an automatic response when evaluating an action. This feeling is strongly based on prior experiences and can color our perception of taking an action again (e.g., remembering a past negative experience like almost having a heart attack when taking the stairs). However, the context does not provide specific details on how different emotions alter decision-making patterns.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:46:53.595672"
+      },
+      {
+        "query": "Does the book compare digital versus physical reward effectiveness?",
+        "answer": "I don't know.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:46:55.370593"
+      },
+      {
+        "query": "How do defaults interact with making habits obvious?",
+        "answer": "Based on the provided context, defaults can make habits more obvious by serving as an initial, automatic step that users can follow without conscious effort, thereby helping them \"get into the groove\" of starting a new behavior. The context states that defaults are powerful because users often stick with them due to being in a hurry, unsure, or lacking strong preferences. While habits require an \"unambiguous cue, unvarying routine, and a meaningful, immediate reward,\" defaults act as a pre-set starting point that can move individuals closer to action, making the initial commitment simpler and reducing the conscious effort needed to begin forming a habit.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:46:55.939969"
+      },
+      {
+        "query": "What is the minimum time between response and reward before the feedback loop breaks?",
+        "answer": "Based on the provided context, there is no specific minimum time mentioned for when the feedback loop breaks. The context emphasizes that rewards occurring **immediately** after the routine are best for strengthening the association, but it does not provide a precise time threshold.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:46:56.568418"
+      },
+      {
+        "query": "What specific environmental cues make good habits obvious in digital spaces?",
+        "answer": "Based on the provided context, specific environmental cues in digital spaces that make good habits obvious are not explicitly detailed. However, the context does state that \"cues, wisely placed, are essential for changes in behavior\" and that within software products, asking for action acts as a cue by increasing attention. It also mentions that habits require \"an unambiguous cue\" and gives the example of feeling tooth film as a cue for brushing teeth. Therefore, while the general principle of using clear, well-placed cues is supported, the context does not list specific examples of such cues in digital environments.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:46:56.875413"
+      },
+      {
+        "query": "Are there examples of habit stacking with inconsistent anchors?",
+        "answer": "I don't know. The provided context does not mention \"habit stacking\" or discuss \"inconsistent anchors\" in relation to habit formation.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:46:58.501306"
+      },
+      {
+        "query": "How far in advance should reminders trigger before deadlines for ADHD users?",
+        "answer": "I don't know.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:46:58.696002"
+      },
+      {
+        "query": "How do you externalize executive functions through interface design?",
+        "answer": "Based on the provided context, I don't have information on how to externalize executive functions through interface design. The context discusses applying behavioral tactics to interface design, moving from conceptual designs to interface designs with constraints, reviewing interface designs, and turning designs into code, but it does not specifically address externalizing executive functions.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:46:58.762421"
+      },
+      {
+        "query": "What time intervals work for ADHD users feeling in sync with clocks?",
+        "answer": "I don't have information about specific time intervals that work for ADHD users feeling in sync with clocks in the provided context.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:02.246311"
+      },
+      {
+        "query": "What is the working memory capacity for ADHD versus neurotypical users?",
+        "answer": "I don't have information about the specific working memory capacity differences between ADHD and neurotypical users in the provided context. The context only mentions George Miller's general finding that immediate memory can hold \"seven (plus or minus two) numbers or other chunks of information\" for people in general, without making comparisons for specific conditions like ADHD.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:02.954127"
+      },
+      {
+        "query": "How do you design habit stacking when users lack reliable daily anchor habits?",
+        "answer": "Based on the provided context, I don't have information on designing habit stacking specifically for users who lack reliable daily anchor habits. The context discusses habit formation (cue-routine-reward) and the difficulty of stopping habits, but does not address the strategy of habit stacking or how to design it when anchor habits are unreliable.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:03.596856"
+      },
+      {
+        "query": "How do you frame streaks as identity markers rather than performance metrics?",
+        "answer": "I don't have enough context to answer that question. The provided text discusses metrics for measuring impact and user behavior, but it does not address how to frame streaks specifically as identity markers versus performance metrics.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:03.986000"
+      },
+      {
+        "query": "What makes external rewards effective for ADHD brains specifically?",
+        "answer": "Based solely on the provided context, there is no information about what makes external rewards effective for ADHD brains specifically.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:04.539737"
+      },
+      {
+        "query": "Does Clear provide examples of UI or design reinforcing identity implicitly?",
+        "answer": "I don't know.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:05.882289"
+      },
+      {
+        "query": "Are there specific scenarios where the endowment effect was stronger or weaker?",
+        "answer": "I don't know.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:08.207633"
+      },
+      {
+        "query": "What makes the endowment effect stronger or weaker?",
+        "answer": "Based on the provided context, I don't have information on what makes the endowment effect stronger or weaker. The context discusses loss aversion and how losses are more motivating than gains, but it does not specify the factors that influence the strength of the endowment effect.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:12.311304"
+      },
+      {
+        "query": "What is the threshold where choice becomes too much?",
+        "answer": "Based on the context, the threshold where choice becomes overwhelming is illustrated by the jam study: a display with 24 options attracted more customers but resulted in only 3% purchasing, while a display with 6 options attracted fewer customers but led to 30% purchasing. This suggests that having too many options can paralyze decision-making and reduce satisfaction. Therefore, the context implies that offering fewer options (like 6 instead of 24) is more effective for facilitating choices and user satisfaction.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:13.627355"
+      },
+      {
+        "query": "What makes temptation bundling effective and when does it fail?",
+        "answer": "Based on the provided context, temptation bundling is effective when it involves bundling something people really like (e.g., reading a book) with something they are less motivated to do (e.g., exercising). The effectiveness relies on intentional, voluntary bundling\u2014offering people the option to engage in both activities simultaneously. It fails when it attempts to hold something people love \"hostage\" to something they dislike, implying forced or coercive bundling is not effective.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:15.424600"
+      },
+      {
+        "query": "Does the book describe techniques to avoid the decoy effect while presenting options?",
+        "answer": "I don't know.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:15.897903"
+      },
+      {
+        "query": "What makes a cue noticeable enough to trigger the habit loop in a digital interface?",
+        "answer": "Based solely on the provided context, a cue in a digital interface must be noticeable enough to trigger the habit loop by being directly tied to another event, occurring at a specific time, or being wisely constructed to avoid ambiguity and competition from other cues. The context states that external triggers can be just as effective as internal ones \"if wisely constructed.\"",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:15.973223"
+      },
+      {
+        "query": "Does Ariely provide examples of expectation priming duration from his experiments?",
+        "answer": "Based on the provided context, I do not have information about Dan Ariely providing examples of \"expectation priming duration\" from his experiments. The context mentions Ariely in relation to a story about \"reward substitution\" and lists his book *Predictably Irrational*, but it does not discuss examples of expectation priming duration from his work.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:16.446404"
+      },
+      {
+        "query": "How do you design for task switching when executive function is impaired?",
+        "answer": "Based on the provided context, I don't have information on designing for task switching when executive function is impaired. The context discusses behavior change models, barriers to action like procrastination and forgetfulness, and strategies such as highlighting benefits and minimizing costs, but it does not specifically address task switching or impaired executive function.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:17.427071"
+      },
+      {
+        "query": "Does Barkley provide specific visual time representations that worked for ADHD users?",
+        "answer": "I don't know.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:18.877842"
+      },
+      {
+        "query": "How do you measure if a habit is easy enough according to the third law?",
+        "answer": "I don't know.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:20.023576"
+      },
+      {
+        "query": "Are there examples of preserving social norms with optional payment systems?",
+        "answer": "Based solely on the provided context, there is no information about examples of preserving social norms with optional payment systems.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:22.368859"
+      },
+      {
+        "query": "How do you automate habits without removing the sense of agency?",
+        "answer": "Based on the provided context, I don't have information on how to automate habits without removing the sense of agency. The context discusses habits, automation, and conscious choice separately, but does not address combining them in a way that preserves agency.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:25.646417"
+      },
+      {
+        "query": "Does the book provide specific time intervals for ADHD users syncing with clocks?",
+        "answer": "I don't know.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:25.852625"
+      },
+      {
+        "query": "How do you avoid the decoy effect while still presenting options?",
+        "answer": "Based on the provided context, I don't have information on how to avoid the decoy effect while presenting options. The context discusses cognitive quirks, user testing, defining target actions, and behavioral strategies like cheating and building habits, but it does not mention the decoy effect.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:26.298578"
+      },
+      {
+        "query": "How long do expectation effects persist after initial priming?",
+        "answer": "Based on the provided context, there is no information about how long expectation effects persist after initial priming.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:30.424500"
+      },
+      {
+        "query": "What priming techniques work in digital interfaces versus physical spaces?",
+        "answer": "Based on the provided context, I don't have information about specific priming techniques that work in digital interfaces versus physical spaces. The context mentions priming in a general sense but does not detail techniques for different environments.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:30.723338"
+      },
+      {
+        "query": "How does emotion dysregulation affect interaction with error messages?",
+        "answer": "I don't know. The provided context does not contain information about how emotion dysregulation affects interaction with error messages.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:30.824232"
+      },
+      {
+        "query": "Do digital rewards like checkmarks and animations satisfy the same as physical rewards?",
+        "answer": "Based on the provided context, I cannot answer whether digital rewards like checkmarks and animations satisfy the same as physical rewards. The context discusses rewards in general (monetary, social, intrinsic) and their motivational properties, but does not specifically compare digital and physical rewards.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:34.927889"
+      },
+      {
+        "query": "What happens when the anchor habit is inconsistent or varies in timing?",
+        "answer": "Based on the provided context, I don't have information about what happens when an anchor habit is inconsistent or varies in timing. The context discusses cues and triggers for habits, mentioning they can be tied to another event or at a specific time of day, but does not address the consequences of inconsistency or variation in timing.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:38.053328"
+      },
+      {
+        "query": "How long do plateaus typically last for different habit types?",
+        "answer": "I don't know. The provided context does not contain information about the typical duration of plateaus for different habit types.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:40.354857"
+      },
+      {
+        "query": "Does the book specify how many habits in a stack before fragility occurs?",
+        "answer": "No, the provided context does not specify a number of habits in a stack before fragility occurs. The concept of \"habit stacks\" or a numerical limit for them is not mentioned in the excerpts.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:40.914639"
+      },
+      {
+        "query": "Should you hide metrics during plateaus or make the plateau itself visible?",
+        "answer": "I don't know. The provided context does not contain information about whether to hide metrics during plateaus or make plateaus visible.",
+        "book": "designingforbehaviorchangeapplyingpsychologyandbehavioraleconomics.embeddings",
+        "timestamp": "2026-02-14T00:47:43.348858"
+      }
+    ],
+    "microinteractions": [
+      {
+        "query": "What makes variable rewards effective and when do they become manipulative?",
+        "answer": "Based on the provided context, variable rewards are effective because they provide intermittent reinforcement of behavior, which is engaging due to its unpredictability and the promise of a rare big reward, as seen with slot machines. They become manipulative when the feedback is disproportionate to the actual outcome (e.g., making small wins seem big with flashing lights and sounds) and when the rules are rigged in the system's favor through weighted randomness that the user is unaware of, leading them to gamble more based on \"near wins\" that activate brain areas associated with winning.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:46:52.691751"
+      },
+      {
+        "query": "Are there examples of progress bars that motivated versus demotivated users?",
+        "answer": "I don't know. The provided context does not contain information about examples of progress bars that motivated versus demotivated users.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:46:55.664641"
+      },
+      {
+        "query": "How does mood state specifically alter decision-making speed and quality?",
+        "answer": "I don't know.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:46:57.385392"
+      },
+      {
+        "query": "What mapping patterns work best for timeline controls and why?",
+        "answer": "Based on the provided context, I don't have information about mapping patterns for timeline controls.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:46:57.561118"
+      },
+      {
+        "query": "How many choices on a page trigger decision paralysis?",
+        "answer": "I don't know. The provided context discusses the effects of abundant choice and the importance of meaningful choices, but it does not specify a particular number of choices that triggers decision paralysis.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:46:57.795118"
+      },
+      {
+        "query": "What percentage of decisions are made unconsciously versus consciously?",
+        "answer": "I don't know.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:46:57.847342"
+      },
+      {
+        "query": "Are there examples of when variable rewards became manipulative in practice?",
+        "answer": "Yes, based on the provided context, slot machines are an example where variable rewards became manipulative in practice. The context states that slot machines use feedback to provide \"intermittent reinforcement of behavior,\" with unpredictable payouts and the rare promise of a big jackpot to keep people playing. This design is described as \"insidious\" and specifically crafted to take money from people. Additionally, \"near wins\" are highlighted as activating brain areas associated with winning even when the player loses, making people want to gamble more.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:46:58.289414"
+      },
+      {
+        "query": "Are there examples showing what makes navigation feel automatic versus requiring thought?",
+        "answer": "Based on the provided context, there is no information about specific examples that show what makes navigation feel automatic versus requiring thought.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:46:59.390025"
+      },
+      {
+        "query": "Does the book describe how to measure or test for erroneous mental models?",
+        "answer": "I don't know.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:46:59.463015"
+      },
+      {
+        "query": "How do you measure if something makes users think too much?",
+        "answer": "Based on the provided context, I don't have information on how to measure if something makes users think too much.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:01.157852"
+      },
+      {
+        "query": "Are specific mapping patterns recommended for time-based controls?",
+        "answer": "Based strictly on the provided context, there is no information about specific mapping patterns for time-based controls.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:01.616890"
+      },
+      {
+        "query": "What makes navigation require thinking versus being automatic?",
+        "answer": "Based on the provided context, I don't have information that directly explains what makes navigation require thinking versus being automatic. The context discusses microinteractions, attention, algorithms, and decision-making, but does not address the specific factors that differentiate conscious navigation from automatic navigation.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:02.581837"
+      },
+      {
+        "query": "How do you test for cognitive friction in usability testing?",
+        "answer": "Based on the provided context, I don't have information on how to test for cognitive friction in usability testing. The context discusses testing microinteractions, focusing on aspects like satisfaction, difficulty, confusion, timing, flow, and the number of clicks/taps, but it does not specifically mention the term \"cognitive friction\" or a method for testing it.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:02.928729"
+      },
+      {
+        "query": "Does Krug provide specific numbers for choice thresholds that trigger paralysis?",
+        "answer": "I don't know.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:05.697171"
+      },
+      {
+        "query": "How do you design for recognition without accidentally creating false memories?",
+        "answer": "Based on the provided context, there is no information on designing for recognition while avoiding the creation of false memories.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:06.084890"
+      },
+      {
+        "query": "What is the relationship between attention span and task complexity?",
+        "answer": "Based on the provided context, there is no information about the relationship between attention span and task complexity.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:06.780740"
+      },
+      {
+        "query": "How do you design for satisficing users versus maximizing users?",
+        "answer": "Based on the provided context, there is no information about designing for satisficing users versus maximizing users.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:08.496698"
+      },
+      {
+        "query": "What is the exact capacity of working memory in items and duration?",
+        "answer": "I don't know.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:09.982337"
+      },
+      {
+        "query": "Does the book specify when instructions helped versus harmed in actual tests?",
+        "answer": "Based on the provided context, the book does not specify when instructions helped versus harmed in actual tests.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:10.560952"
+      },
+      {
+        "query": "How do users decide what is the first reasonable option when satisficing?",
+        "answer": "I don't know.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:10.877862"
+      },
+      {
+        "query": "Does the book specify optimal animation duration or speed for attention?",
+        "answer": "Based on the provided context, the book does not specify an optimal animation duration or speed specifically for focusing attention. However, it provides general guidelines for animation:\n\n1.  **Timing Rule:** Scott's rule states to make any animation half as long as you think it should be, and then possibly halve it again.\n2.  **General Duration:** Animations should be \"one second in duration, and usually a fraction of a second.\"\n3.  **Purpose for Attention:** One listed reason for using animation is \"Focusing attention. As an item changes value, an animation can make that change more obvious.\"\n4.  **Characteristics:** Animations should be \"Fast\" (not delaying the activity) and \"Purposeful\" (not just eye candy).\n\nWhile animation is recommended as a method to focus attention, the context does not provide a specific optimal speed or duration for that particular purpose.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:12.899241"
+      },
+      {
+        "query": "What is the threshold where instructions become necessary versus harmful?",
+        "answer": "Based on the provided context, the threshold for when instructions become necessary versus harmful is not explicitly defined. However, the context advises to \"Keep copy short\" and \"Never use instructional text where a label will suffice.\" This suggests a principle of minimizing instructions, favoring simpler interface elements like labels whenever possible.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:13.959021"
+      },
+      {
+        "query": "How long does it take for users to habituate to repetitive notifications?",
+        "answer": "I don't know.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:16.088345"
+      },
+      {
+        "query": "How do people update their mental models after encountering errors?",
+        "answer": "Based on the provided context, there is no information about how people update their mental models after encountering errors.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:19.991714"
+      },
+      {
+        "query": "What makes progress bars motivating versus anxiety-inducing?",
+        "answer": "Based on the provided context, there is no information about what makes progress bars motivating versus anxiety-inducing.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:25.041736"
+      },
+      {
+        "query": "How do you measure if something is providing intrinsic versus extrinsic motivation?",
+        "answer": "I don't know.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:27.050884"
+      },
+      {
+        "query": "What is the maximum acceptable feedback delay before users perceive unresponsiveness?",
+        "answer": "I don't know.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:28.890589"
+      },
+      {
+        "query": "What percentage range is given for unconscious versus conscious decisions?",
+        "answer": "I don't know.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:30.833632"
+      },
+      {
+        "query": "Does the book provide any examples or case studies of notification habituation timeframes?",
+        "answer": "Based strictly on the provided context, I do not have information about examples or case studies of notification habituation timeframes. The context discusses system trigger rules, loops, and data collection, but does not mention \"habituation\" or specific timeframes for user adaptation to notifications.",
+        "book": "microinteractions.embeddings",
+        "timestamp": "2026-02-14T00:47:45.097794"
+      }
+    ],
+    "takingchargeofadultadhd": [
+      {
+        "query": "Does the book provide examples of minimum viable actions that worked for exercise or sleep?",
+        "answer": "Based on the provided context, there is no mention of examples of minimum viable actions specifically for exercise or sleep. The context discusses strategies for managing ADHD, such as externalizing information and using rules for success, but does not detail any minimum viable actions for exercise or sleep routines.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:46:57.506718"
+      },
+      {
+        "query": "Should you hide metrics during plateaus or make the plateau itself visible?",
+        "answer": "I don't know. The provided context does not contain information about whether to hide or make visible metrics during performance plateaus.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:46:58.496433"
+      },
+      {
+        "query": "How long do plateaus typically last for different habit types?",
+        "answer": "Based solely on the provided context, there is no information about how long plateaus typically last for different habit types.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:00.114474"
+      },
+      {
+        "query": "Can you design UI that reinforces identity without explicitly stating it?",
+        "answer": "Based on the provided context, there is no information about designing UI (user interface) or how it relates to reinforcing identity. The context discusses working memory, self-regulation, emotional control, and related cognitive functions in adults with ADHD. Therefore, I don't know the answer to your question.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:01.727699"
+      },
+      {
+        "query": "Are there examples of habit stacking with inconsistent anchors?",
+        "answer": "I don't have information in the provided context about examples of habit stacking with inconsistent anchors.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:02.028399"
+      },
+      {
+        "query": "How do you frame streaks as identity markers rather than performance metrics?",
+        "answer": "Based on the provided context, there is no information on framing streaks as identity markers rather than performance metrics.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:05.641842"
+      },
+      {
+        "query": "How do you design habit stacking when users lack reliable daily anchor habits?",
+        "answer": "Based on the provided context, I don't have information on how to design habit stacking when users lack reliable daily anchor habits. The context discusses rules for everyday success, breaking tasks into smaller steps, and using external reminders, but it does not address the specific strategy of habit stacking or how to design it without anchor habits.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:06.780192"
+      },
+      {
+        "query": "Does Clear provide examples of UI or design reinforcing identity implicitly?",
+        "answer": "I don't know.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:07.301743"
+      },
+      {
+        "query": "Are there examples of first-time experiences that led to high retention?",
+        "answer": "Based on the provided context, there is no specific mention of examples of first-time experiences that led to high retention. The context discusses challenges with memory and learning for adults with ADHD, such as having a limited store of past experiences and difficulties with nonverbal working memory, but it does not provide concrete examples of first-time experiences resulting in high retention.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:07.699465"
+      },
+      {
+        "query": "How do you create compelling reasons to return without using dark patterns?",
+        "answer": "Based on the provided context, there is no information on how to create compelling reasons to return without using dark patterns.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:09.755482"
+      },
+      {
+        "query": "How do you measure whether users are motivated enough to act on feedback?",
+        "answer": "Based on the provided context, there is no information on how to measure whether users are motivated enough to act on feedback.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:11.241717"
+      },
+      {
+        "query": "What makes temptation bundling effective and when does it fail?",
+        "answer": "Based on the provided context, there is no information about \"temptation bundling,\" what makes it effective, or when it fails. I don't know.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:11.242164"
+      },
+      {
+        "query": "What makes a cue noticeable enough to trigger the habit loop in a digital interface?",
+        "answer": "Based on the provided context, there is no information about what makes a cue noticeable enough to trigger a habit loop in a digital interface. The context discusses physical, external cues like pictures, stop signs, and lists, but does not address digital interfaces.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:13.358167"
+      },
+      {
+        "query": "How do you identify when to simplify the behavior versus redesign the environment?",
+        "answer": "Based on the provided context, the decision to simplify behavior versus redesign the environment is guided by the principle of making problems external, physical, and manual. Specifically, you should redesign the environment when you need to externalize information that is usually held in the mind, putting key pieces of information into a physical form where the problem exists. This approach helps when you struggle with holding relevant facts in your mind for problem-solving, weighing pros and cons, or analyzing approaches. The context emphasizes modifying your environment to assist with the time, timing, and timeliness of behavior, and using physical cues in plain view in problem situations.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:16.185317"
+      },
+      {
+        "query": "Does the book specify how many habits in a stack before fragility occurs?",
+        "answer": "No, the provided context does not specify a number of habits in a stack before fragility occurs.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:17.988653"
+      },
+      {
+        "query": "How do you measure if feedback is actionable versus just informative?",
+        "answer": "Based on the provided context, there is no information about measuring whether feedback is actionable versus just informative. The context discusses the feedback conference in an ADHD evaluation, detailing what the professional will do and the diagnostic criteria, but it does not address the distinction between actionable and informative feedback or how to measure it. Therefore, I don't know.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:19.858015"
+      },
+      {
+        "query": "How do you automate habits without removing the sense of agency?",
+        "answer": "Based on the provided context, I don't have information on how to automate habits without removing the sense of agency. The context discusses strategies like using external tools (timers, journals, alarms), creating budgets, and performing simple actions to manage impulses, but it does not address the specific balance between automation and personal agency.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:20.373429"
+      },
+      {
+        "query": "What is the minimum viable action for common behaviors like exercise or sleep routines?",
+        "answer": "Based on the provided context, there is no specific information about the minimum viable action for common behaviors like exercise or sleep routines.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:22.938612"
+      },
+      {
+        "query": "What happens when the anchor habit is inconsistent or varies in timing?",
+        "answer": "Based on the provided context, there is no information about what happens when an anchor habit is inconsistent or varies in timing. The context discusses concepts related to self-control, time management, and ADHD, but does not address anchor habits specifically. Therefore, I don't know.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:25.130992"
+      },
+      {
+        "query": "What is the minimum time between response and reward before the feedback loop breaks?",
+        "answer": "Based on the provided context, there is no specific minimum time mentioned between response and reward before the feedback loop breaks. The context discusses the importance of self-control in bridging time gaps and the challenges of delayed outcomes, but it does not specify a minimum threshold.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:26.458609"
+      },
+      {
+        "query": "Do digital rewards like checkmarks and animations satisfy the same as physical rewards?",
+        "answer": "Based on the provided context, there is no information comparing digital rewards like checkmarks and animations to physical rewards. Therefore, I don't know.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:26.533255"
+      },
+      {
+        "query": "How do you measure if a habit is easy enough according to the third law?",
+        "answer": "The provided context does not contain information about a \"third law\" or a method for measuring if a habit is \"easy enough\" according to it. Therefore, I don't know.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:28.813983"
+      },
+      {
+        "query": "What specific environmental cues make good habits obvious in digital spaces?",
+        "answer": "Based on the provided context, specific environmental cues that make good habits obvious in digital spaces are not mentioned. The context discusses using physical cues like pictures, stop signs, and lists placed on mirrors, refrigerators, or desks, but does not extend this advice to digital environments.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:29.567500"
+      },
+      {
+        "query": "How do defaults interact with making habits obvious?",
+        "answer": "Based on the provided context, there is no information about how defaults interact with making habits obvious.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:31.923430"
+      },
+      {
+        "query": "What specific friction increases work for breaking bad habits without causing abandonment?",
+        "answer": "Based on the provided context, there is no specific information about \"friction\" or techniques for increasing work to break bad habits without causing abandonment. The context discusses strategies like using physical prompts (cards, lists, signs) and externalizing rules, but does not address the concept of strategically increasing friction for habit change.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:32.228079"
+      },
+      {
+        "query": "Does the book compare digital versus physical reward effectiveness?",
+        "answer": "Based strictly on the provided context, there is no direct comparison between digital versus physical reward effectiveness mentioned in the excerpts. The context discusses incentives and motivation in general terms but does not specify a comparison between digital and physical rewards. Therefore, I don't know if the book makes such a comparison.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:36.639459"
+      },
+      {
+        "query": "Are there case studies of friction levels that broke bad habits without causing dropout?",
+        "answer": "Based strictly on the provided context, there is no information about case studies involving friction levels for breaking bad habits without causing dropout. The context discusses ADHD symptoms and management strategies, but does not mention specific case studies on this topic.\n\nTherefore, I do not know.",
+        "book": "takingchargeofadultadhd.embeddings",
+        "timestamp": "2026-02-14T00:47:40.712838"
+      }
+    ],
+    "irrational": [
+      {
+        "query": "How do reversible choices affect satisfaction compared to final ones?",
+        "answer": "Based on the provided context, there is no information about reversible choices and their effect on satisfaction compared to final ones.",
+        "book": "irrational.embeddings",
+        "timestamp": "2026-02-14T00:46:59.031604"
+      },
+      {
+        "query": "Does Schwartz provide a specific optimal number of choices for any decision types?",
+        "answer": "I don't know.",
+        "book": "irrational.embeddings",
+        "timestamp": "2026-02-14T00:47:01.608904"
+      },
+      {
+        "query": "What is good enough and how do you communicate that standard?",
+        "answer": "Based on the provided context, there is no information about what constitutes \"good enough\" or how to communicate that standard.",
+        "book": "irrational.embeddings",
+        "timestamp": "2026-02-14T00:47:05.768800"
+      },
+      {
+        "query": "What is the actual optimal number of choices for different decision types?",
+        "answer": "Based strictly on the provided context, there is no information about the optimal number of choices for different decision types. The context discusses how the presence or absence of a decoy option influences choices, but it does not specify optimal numbers for decision types.",
+        "book": "irrational.embeddings",
+        "timestamp": "2026-02-14T00:47:07.595430"
+      },
+      {
+        "query": "What specific defaults reduce choice overload most effectively?",
+        "answer": "Based on the provided context, I do not have information on specific defaults that reduce choice overload most effectively. The context mentions research on defaults in general (e.g., \"Do Defaults Save Lives?\") but does not detail which defaults are most effective for alleviating choice overload.",
+        "book": "irrational.embeddings",
+        "timestamp": "2026-02-14T00:47:09.462577"
+      },
+      {
+        "query": "How do you reduce anticipated regret through design?",
+        "answer": "Based on the provided context, anticipated regret can be reduced through design by having restaurant owners ask customers to write orders privately or give orders quietly to waiters. This prevents customers from being influenced by others' choices, which can lead to regrettable mistakes.",
+        "book": "irrational.embeddings",
+        "timestamp": "2026-02-14T00:47:14.298214"
+      }
+    ],
+    "paradoxofchoice": [
+      {
+        "query": "How long do expectation effects persist after initial priming?",
+        "answer": "I don't know.",
+        "book": "paradoxofchoice.embeddings",
+        "timestamp": "2026-02-14T00:47:15.102597"
+      },
+      {
+        "query": "What priming techniques work in digital interfaces versus physical spaces?",
+        "answer": "I don't have information about priming techniques in digital interfaces versus physical spaces in the provided context.",
+        "book": "paradoxofchoice.embeddings",
+        "timestamp": "2026-02-14T00:47:18.688954"
+      },
+      {
+        "query": "Are there specific scenarios where the endowment effect was stronger or weaker?",
+        "answer": "Based on the provided context, there are specific scenarios where the endowment effect was stronger or weaker. It was stronger when options were already attached to a car (part of the endowment), leading people to keep more options because giving them up felt like a loss. Conversely, the effect was weaker when options were not initially attached, as adding them was perceived as a gain, and people chose fewer options.",
+        "book": "paradoxofchoice.embeddings",
+        "timestamp": "2026-02-14T00:47:20.055279"
+      },
+      {
+        "query": "Does the book describe techniques to avoid the decoy effect while presenting options?",
+        "answer": "I don't know.",
+        "book": "paradoxofchoice.embeddings",
+        "timestamp": "2026-02-14T00:47:20.427712"
+      },
+      {
+        "query": "What makes the endowment effect stronger or weaker?",
+        "answer": "Based on the provided context, the endowment effect is influenced by whether an item is already part of one's endowment. When options are already attached to a product (like a car), they become part of the endowment, and giving them up feels like a loss. Conversely, when options are not attached, choosing them is perceived as a gain. The effect is stronger when people own an item, as seen in the study where participants demanded 30% more to sell a mug after owning it compared to before owning it. Additionally, the endowment effect operates even before a purchase is finalized, affecting decisions based on perceived losses versus gains.",
+        "book": "paradoxofchoice.embeddings",
+        "timestamp": "2026-02-14T00:47:21.170629"
+      },
+      {
+        "query": "How do different emotions specifically alter decision-making patterns?",
+        "answer": "Based on the provided context, negative emotions narrow our focus, causing us to home in on only one or two aspects of a decision while potentially ignoring important ones, and distract us by making us focus on the emotion itself. This impairs decision-making, especially as stakes rise. Conversely, positive emotion has the opposite effect: it helps us think better, consider more possibilities, be open to otherwise unconsidered factors, and see subtle connections between pieces of information.",
+        "book": "paradoxofchoice.embeddings",
+        "timestamp": "2026-02-14T00:47:23.909169"
+      },
+      {
+        "query": "Does Ariely provide examples of expectation priming duration from his experiments?",
+        "answer": "I don't know.",
+        "book": "paradoxofchoice.embeddings",
+        "timestamp": "2026-02-14T00:47:25.974537"
+      },
+      {
+        "query": "Are there examples of preserving social norms with optional payment systems?",
+        "answer": "Based on the provided context, there is no mention of examples specifically related to preserving social norms with optional payment systems. The context discusses constraints on choice, second-order decisions like following rules, and market freedom, but does not address optional payment systems in this regard. Therefore, I don't know.",
+        "book": "paradoxofchoice.embeddings",
+        "timestamp": "2026-02-14T00:47:26.719869"
+      },
+      {
+        "query": "How do you avoid the decoy effect while still presenting options?",
+        "answer": "Based on the provided context, there is no information about the \"decoy effect\" or strategies to avoid it while presenting options. The context discusses the psychological costs of keeping options open, opportunity costs, and the difficulty of decisions involving trade-offs, but does not address the specific concept of a decoy effect.\n\nTherefore, I don't know how to avoid the decoy effect based on the given context.",
+        "book": "paradoxofchoice.embeddings",
+        "timestamp": "2026-02-14T00:47:28.890456"
+      },
+      {
+        "query": "What is the threshold where choice becomes too much?",
+        "answer": "Based on the provided context, there is no specific numerical threshold mentioned for when choice becomes too much. The context discusses that the negative effects of choice escalate as the number of choices increases and are subtle and cumulative, but it does not define a precise point or quantity.",
+        "book": "paradoxofchoice.embeddings",
+        "timestamp": "2026-02-14T00:47:29.369264"
+      }
+    ]
+  }
+}
\ No newline at end of file
diff --git a/scheduler/daemon.py b/scheduler/daemon.py
index c3dc659..03d0acd 100644
--- a/scheduler/daemon.py
+++ b/scheduler/daemon.py
@@ -33,6 +33,7 @@ def check_medication_reminders():
     """Check for medications due now and send notifications."""
     try:
         from datetime import date as date_type
+
         meds = postgres.select("medications", where={"active": True})
 
         # Group by user so we only look up timezone once per user
@@ -68,8 +69,14 @@ def check_medication_reminders():
                     start = med.get("start_date")
                     interval = med.get("interval_days")
                     if start and interval:
-                        start_d = start if isinstance(start, date_type) else datetime.strptime(str(start), "%Y-%m-%d").date()
-                        if (today - start_d).days < 0 or (today - start_d).days % interval != 0:
+                        start_d = (
+                            start
+                            if isinstance(start, date_type)
+                            else datetime.strptime(str(start), "%Y-%m-%d").date()
+                        )
+                        if (today - start_d).days < 0 or (
+                            today - start_d
+                        ).days % interval != 0:
                             continue
                     else:
                         continue
@@ -80,7 +87,9 @@ def check_medication_reminders():
                     continue
 
                 # Already taken today? Check by created_at date
-                logs = postgres.select("med_logs", where={"medication_id": med["id"], "action": "taken"})
+                logs = postgres.select(
+                    "med_logs", where={"medication_id": med["id"], "action": "taken"}
+                )
                 already_taken = any(
                     log.get("scheduled_time") == current_time
                     and str(log.get("created_at", ""))[:10] == today_str
@@ -91,8 +100,10 @@ def check_medication_reminders():
 
                 user_settings = notifications.getNotificationSettings(user_uuid)
                 if user_settings:
-                    msg = f"Time to take {med['name']} ({med['dosage']} {med['unit']})"
-                    notifications._sendToEnabledChannels(user_settings, msg, user_uuid=user_uuid)
+                    msg = f"Time to take {med['name']} ({med['dosage']} {med['unit']}) · {current_time}"
+                    notifications._sendToEnabledChannels(
+                        user_settings, msg, user_uuid=user_uuid
+                    )
     except Exception as e:
         logger.error(f"Error checking medication reminders: {e}")
 
@@ -120,7 +131,9 @@ def check_routine_reminders():
             user_settings = notifications.getNotificationSettings(routine["user_uuid"])
             if user_settings:
                 msg = f"Time to start your routine: {routine['name']}"
-                notifications._sendToEnabledChannels(user_settings, msg, user_uuid=routine["user_uuid"])
+                notifications._sendToEnabledChannels(
+                    user_settings, msg, user_uuid=routine["user_uuid"]
+                )
     except Exception as e:
         logger.error(f"Error checking routine reminders: {e}")
 
@@ -135,7 +148,9 @@ def check_refills():
                 user_settings = notifications.getNotificationSettings(med["user_uuid"])
                 if user_settings:
                     msg = f"Low on {med['name']}: only {qty} doses remaining. Time to refill!"
-                    notifications._sendToEnabledChannels(user_settings, msg, user_uuid=med["user_uuid"])
+                    notifications._sendToEnabledChannels(
+                        user_settings, msg, user_uuid=med["user_uuid"]
+                    )
     except Exception as e:
         logger.error(f"Error checking refills: {e}")