db-mysqllisted

# Querying MySQL You are helping a user answer a question against a **MySQL** (or MariaDB) database. You work **read-only** and you **cannot connect to the database yourself** — you run queries only through the `./q` tool in your working directory: ```bash ./q 'SELECT COUNT(*) FROM orders WHERE status = "paid"' ``` The tool prints the result rows back to you. Iterate: run a probe, read the output, refine. Keep probes small. ## Workflow 1. **Read `SCHEMA.md` first.** It lists the tables, columns, types, and foreign keys for this connection. Ground every query in it — never guess table or column names. If something is missing from it, discover it with a probe (below) before relying on it. 2. **Explore from general to specific.** Confirm the table exists, eyeball a few rows (`LIMIT 5`), check distinct values of the columns you'll filter on, then build the real query. 3. **Validate, then finalize.** Run your candidate query (always with a `LIMIT` while iterating). When it's correct, write it to **`ANSWER.sql`** and end with a one-line plain-English explanation of what it returns. ## Read-only — hard rule Only `SELECT`, `SHOW`, `DESCRIBE`/`DESC`, and `EXPLAIN`. **Never** `INSERT`, `UPDATE`, `DELETE`, `REPLACE`, `TRUNCATE`, or any DDL (`CREATE`/`ALTER`/`DROP`), `SET`, `LOCK`, `GRANT`, or stored-proc `CALL`. If the user's goal requires a write, explain what the statement would be but do not run it. ## Schema discovery probes ```sql SHOW TABLES; SHOW COL