Alternative to Huckleberry: a baby sleep and feeding log with ChatGPT

Set up a baby log for me in databasedmcp, in one go.

1. Create a database called baby_log, described as "Baby log for Ilo: sleep, feeds and diapers".
2. Create three tables:
   - sleeps: id, started_at (date and time, defaults to now), ended_at (NULL while she's asleep), kind (nap or night), notes.
   - feeds: id, at (date and time, defaults to now), kind (breast, bottle or solids), side (left, right or both, for breast feeds), amount_ml, duration_min, notes.
   - diapers: id, at (date and time, defaults to now), kind (wet, dirty or both), notes.
3. Write the database a guide that says what each table is for, that times are stored in UTC and shown to me in my local time, that only one sleep is open at a time, and the queries for "last feeds", "hours of sleep per day this week" and "feeds and diapers per day".

From then on: when I say she fell asleep, start a sleep; when she wakes, set ended_at on the open one. Log every feed and diaper I mention. Keep replies short. Don't give medical advice; if I ask about something worrying, tell me to ask a health professional. Show me the tables when you're finished.

CREATE TABLE sleeps (
  id INTEGER PRIMARY KEY,
  started_at TEXT NOT NULL DEFAULT CURRENT_TIMESTAMP,
  ended_at TEXT,
  kind TEXT CHECK (kind IN ('nap', 'night')),
  notes TEXT
);

CREATE TABLE feeds (
  id INTEGER PRIMARY KEY,
  at TEXT NOT NULL DEFAULT CURRENT_TIMESTAMP,
  kind TEXT NOT NULL CHECK (kind IN ('breast', 'bottle', 'solids')),
  side TEXT CHECK (side IN ('left', 'right', 'both')),
  amount_ml INTEGER,
  duration_min INTEGER,
  notes TEXT
);

CREATE TABLE diapers (
  id INTEGER PRIMARY KEY,
  at TEXT NOT NULL DEFAULT CURRENT_TIMESTAMP,
  kind TEXT NOT NULL CHECK (kind IN ('wet', 'dirty', 'both')),
  notes TEXT
);

# baby_log: Ilo's sleep, feeds and diapers

- **sleeps**: one row per sleep. `ended_at` is NULL while she's asleep; only one sleep is open at a time. kind: nap or night.
- **feeds**: breast (with side and minutes), bottle (with ml) or solids.
- **diapers**: wet, dirty or both.

Times are UTC (datetime('now')); show them in my local time. This is a log, not advice: for anything worrying, say to ask a health professional.

Last feeds: `SELECT at, kind, side, amount_ml, duration_min FROM feeds ORDER BY at DESC LIMIT 3`

Sleep per day: `SELECT date(started_at) AS day, count(*) AS sleeps, round(sum(julianday(ended_at) - julianday(started_at)) * 24, 1) AS hours FROM sleeps WHERE ended_at IS NOT NULL AND started_at >= date('now', '-7 days') GROUP BY day ORDER BY day`

INSERT INTO feeds (at, kind, side, duration_min) VALUES (datetime('now', ?), ?, ?, ?)
-- params: ["-60 minutes","breast","left",15]

INSERT INTO sleeps (started_at, kind) VALUES (datetime('now', ?), ?)
-- params: ["-40 minutes","nap"]

UPDATE sleeps SET ended_at = datetime('now') WHERE ended_at IS NULL

INSERT INTO diapers (kind) VALUES (?)
-- params: ["wet"]

INSERT INTO feeds (kind, amount_ml) VALUES (?, ?)
-- params: ["bottle",120]

SELECT at, kind, side, amount_ml, duration_min, round((julianday('now') - julianday(at)) * 24, 1) AS hours_ago FROM feeds ORDER BY at DESC LIMIT 3

SELECT date(started_at) AS day, count(*) AS sleeps, round(sum(julianday(ended_at) - julianday(started_at)) * 24, 1) AS hours FROM sleeps WHERE ended_at IS NOT NULL AND started_at >= date('now', '-7 days') GROUP BY day ORDER BY day

WITH f AS (SELECT date(at) AS day, count(*) AS feeds FROM feeds GROUP BY day), d AS (SELECT date(at) AS day, count(*) AS diapers FROM diapers GROUP BY day) SELECT f.day, f.feeds, coalesce(d.diapers, 0) AS diapers FROM f LEFT JOIN d ON d.day = f.day WHERE f.day >= date('now', '-7 days') ORDER BY f.day DESC