drive-recommendations

Drive Recommendations

When to use

Activate this skill when the user asks anything along the lines of:

• "Any advice on how I drive?"
• "How can I get more range?"
• "Give me recommendations on my drives."
• "What should I change in my driving habits?"
• "Tell me one thing to improve in how I drive."

If the user wants a full breakdown (median Wh/km, outliers list, baseline analysis) instead of focused advice, prefer the drive-efficiency-coach skill.

Prerequisites

You need a TeslaMate car_id. If unknown:

1. Call teslamate_get_cars to list available cars.
2. If exactly one is returned, use it silently.
3. If multiple, ask the user which one before continuing.

Workflow

1. Pull the dataset (current + previous period)

Call teslamate_get_car_drives twice:

• Current period — default: 30 days ago → now (RFC3339). If the user specifies a different period, use that instead.
• Previous period — same duration, immediately preceding the current window. For a 30-day current window, that means days -60 → -30.

Both calls use:

• car_id: target car
• min_distance: 5 (filters out parking-lot moves where Wh/km is meaningless)
• fetch_all: True whenever the window is longer than ~1 month — the TeslaMate API pages at 100 entries.

Always pass both start_date and end_date. Skip the previous-period call only if the user explicitly says "just this month" / "no comparison".

2. Compute per-drive signals

For each drive:

• Distance in km (convert from miles if TeslaMate is configured that way).
• Energy used, from whichever field is present (consumption_kwh, range delta × nominal Wh/km, or SoC delta × usable pack capacity).
• wh_per_km = energy_used_wh / distance_km
• avg_speed_kmh = distance_km / (duration_min / 60)
• Speed band: city (< 50 km/h), mixed (50–90), highway (> 90).
• Trip length: short (< 15 km), medium (15–60), long (> 60).
• Outside temperature if exposed.
• Skip drives where distance_km < 5 or energy_used_wh <= 0.