Home Interactive Response Skill

Purpose

Load this skill when an interactive home message needs a deliberate response
mode or a worked scene, comfort, status, energy, health-check, setup, or
confirmation pattern.

Response Mode Selection

1. React: Emoji-only acknowledgment

   • Use when: The action is simple and self-explanatory
   • Example: User requests scene execution → React with ✅

2. Affirm: Brief confirmation message

   • Use when: Need a short confirmation with key information
   • Example: "Scene 'Movie Night' activated: lights dimmed, blinds closed."

3. Follow-up: Proactive question or observation

   • Use when: Need more information or can offer recommendations
   • Example: "I've set comfort preferences for your bedroom. Should I apply the same settings to the living room?"

4. Answer: Substantive response to a question

   • Use when: User asked a direct question about home status or data
   • Example: "Living room is at 72°F and 45% humidity, both in your comfort range. Bedroom is 70°F."

5. React + Reply: Combined emoji + message

   • Use when: Immediate acknowledgment plus substantive feedback
   • Example: React with ✅ then "Door locked. All windows closed. Security system armed."

Complete Examples

Example 1: Scene Activation (Affirm)

User message: "Activate movie night"

Actions:

1. ha_activate_scene(entity_id="scene.movie_night")
2. memory_store_fact(subject="movie-night-scene", predicate="usage_pattern", content="user triggered movie night scene", permanence="standard", importance=4.0, tags=["scene", "movie-night"])
3. notify(channel="telegram", message="Scene 'Movie Night' activated — lights dimmed to 20%, blinds closed, sound system on.", intent="reply", request_context=...)

Example 2: Comfort Query (Answer)

User message: "What's the temperature in here?"

Actions:

1. ha_list_entities(domain="sensor", area="living_room") to find temperature/humidity sensors
2. ha_get_entity_state(entity_id="sensor.living_room_temperature") returns 72°F
3. ha_get_entity_state(entity_id="sensor.living_room_humidity") returns 48%
4. memory_recall(subject="living-room", predicate="comfort_preference") to get target range
5. notify(channel="telegram", message="Living room is at 72°F and 48% humidity — both in your comfort range. Your target is 71-73°F.", intent="reply", request_context=...)

Example 3: Preference Setting (Affirm)

User message: "I like the bedroom cooler at night, around 68 degrees"

Actions:

1. memory_store_fact(subject="bedroom", predicate="comfort_preference", content="user prefers 68°F in bedroom at night", permanence="stable", importance=7.0, tags=["bedroom", "temperature", "comfort", "night"])
2. ha_call_service(domain="climate", service="set_temperature", target={"area_id": "bedroom"}, data={"temperature": 68})
3. notify(channel="telegram", message="Set bedroom night temperature to 68°F (67-69°F range). I'll adjust the thermostat tonight.", intent="reply", request_context=...)

Example 4: Device Status with Alert (Answer)

User message: "Are all the lights off?"

Actions:

1. ha_list_entities(domain="light") to get all light entities with current state
2. Parse result: living room light (on, brightness 77/255 ≈ 30%), bedroom light (off), kitchen light (off)
3. notify(channel="telegram", message="Living room light is still on at 30% brightness. Other lights are off. Want me to turn it off?", intent="reply", request_context=...)

Example 5: Energy Report (Answer)

User message: "What's my energy usage today?"

Actions:

1. ha_list_entities(domain="sensor") to find energy monitoring sensors (look for entities with energy or kwh in the name)
2. ha_get_statistics(statistic_ids=["sensor.energy_consumption_kwh"], start="<today 00:00 ISO>", end="<now ISO>", period="hour") to get hourly energy totals
3. ha_get_statistics(statistic_ids=["sensor.hvac_energy", "sensor.water_heater_energy"], start="<today 00:00 ISO>", end="<now ISO>", period="day") for the per-device breakdown
4. notify(channel="telegram", message="Today's usage: 18 kWh. Top consumers: HVAC (40%), Water heater (25%), Kitchen appliances (15%). Usage is 8% below your weekly average.", intent="reply", request_context=...)

Example 6: Proactive Device Health Check (Follow-up)

User message: "Run the device health check"

Actions:

1. ha_list_entities() to survey all entities; look for any with state unavailable or unknown
2. ha_list_entities(domain="sensor") to find battery sensors; look for names containing battery
3. ha_get_entity_state(entity_id="sensor.basement_sensor_battery") returns 15%
4. memory_store_fact(subject="basement-sensor", predicate="device_issue", content="battery at 15% — critical; needs replacement soon", permanence="volatile", importance=8.0, tags=["maintenance", "urgent", "battery"])
5. notify(channel="telegram", message="Health check complete. One device needs attention: basement sensor battery is at 15%.", intent="reply", request_context=...)

Example 7: Scene Setup (Affirm + Follow-up)

User message: "Create a bedtime scene that cools the bedroom to 68 and dims all lights"

Actions:

1. ha_list_entities(domain="light") to discover all light entity IDs
2. ha_call_service(domain="climate", service="set_temperature", target={"area_id": "bedroom"}, data={"temperature": 68})
3. ha_call_service(domain="light", service="turn_on", target={"area_id": "all"}, data={"brightness_pct": 10})
4. ha_call_service(domain="scene", service="create", data={"scene_id": "bedtime", "entities": {"climate.bedroom": {"state": "cool", "temperature": 68}, "light.living_room": {"state": "on", "brightness": 26}, "light.bedroom": {"state": "on", "brightness": 26}}})
5. memory_store_fact(subject="bedtime-scene", predicate="scene_preference", content="user created bedtime scene with temperature 68 and dim lighting", permanence="standard", importance=6.0, tags=["scene", "bedtime", "automation"])
6. notify(channel="telegram", message="Bedtime scene created — bedroom cooling to 68°F, all lights dimmed to 10%. Want me to schedule it for 10pm every night?", intent="reply", request_context=...)

Example 8: Destructive Action Confirmation (Follow-up)

User message: "Disable the movie night automation"

Actions:

1. Instead of immediately disabling, ask for confirmation via notify()
2. notify(channel="telegram", message="Disable 'Movie Night' automation? React with ✅ to confirm or ❌ to cancel.", intent="reply", request_context=...)
3. Wait for user confirmation before calling ha_call_service(domain="automation", service="turn_off", target={"entity_id": "automation.movie_night"})