Ask which indexes your table is missing
"Analyze the customers table and suggest indexes to make it run faster." SnoutData reads the table's real structure and its existing indexes, then tells you what is missing and why, with the CREATE INDEX statements ready to run.
The assistant starts from the facts: it describes the table, checks which indexes already exist, and looks at the shape of the data before it recommends anything. You get a ranked list of gaps (a missing status index, a time-based index for created_at, a composite for filters you run together) instead of generic advice that ignores your schema.
Every suggestion comes with its reasoning
Each recommendation says why it helps, roughly how much it is likely to matter, and the exact statement to add it. It also flags the trade-offs, because an index that speeds up reads costs you on writes, and you should decide that with the numbers in front of you.
From advice to migration
The output is runnable SQL, not a lecture. Drop the CREATE INDEX statements into an editor tab, review them, and run the ones that fit your workload. The assistant surfaces the candidates; you stay in control of what actually touches the database.