Primary Key

You might say

AI said to update this record by id. Where does that id come from?

The field in a table used to uniquely identify each rowIf two rows in a users table share an email, an update by email matches both; an update limited to primary key id = 7 identifies the target row. Check the matching count before writing and read the table again to confirm the other row stayed unchanged. A primary key can be one column or a group, and its value is unique and non-null.

Know first

Database

Primary KeyID

Which column uniquely identifies a row

users · before update
id	name	email	city
7	Lin	lin@example.com	Hangzhou
8	Lin Xiao	lin@example.com	Suzhou
9	Chen Yu	chen@example.com	Xiamen

Two accounts in this users table happen to share an email. Their primary keys, id values, differ, so each can identify one row; email is not this table’s primary key.

Why a primary key must be unique

Locate one row, not a batch: A primary key uniquely identifies a row; still inspect the write condition, affected-row count, and read-back result. A repeated field can match multiple rows.

Ordinary fields repeat: Identical names and emails happen all the time; a primary key's values cannot repeat and cannot be empty.

Tables link through it: When an orders table records which user an order belongs to, it stores the users table's primary key, not the name.

You can say this to an AI Agent

Update this user record by primary key, not by email or name. Before updating, tell me how many rows match; if more than one, stop and explain before continuing. After the change, query the record again to confirm.

Learn next

1Database→2Field→3Database Write→

Recommended tool

Web fine-tuning tool · Selector Pro

Select page elements for precise AI edits

Click a button, text, or card, note the change, then copy its exact location to AI.

Open Selector Pro website

You want to update someone's profile by email in a user table that has two identical emails. What is the problem?

A. Update any one of the rows; the impact is smallB. Delete all duplicate-email records first and keep oneC. This table has repeated emails; use its defined primary key id to identify the target row