Verify a SQL aggregation query
You asked an AI assistant for a SQL query to count active users per plan in the last 30 days, and it gave you a confident-looking query with a JOIN and a GROUP BY. Before you paste it into a dashboard that execs will read, what quick check do you run to confirm the numbers are trustworthy?
Implement
reconcile_plan_totals(plan_counts: list[int], independent_total: int) → strExamples
in
[[120,80,50],250]out"match"in
[[120,95,50],250]out"fan_out:15"What a strong answer looks like
Treat the AI’s output as a draft to verify, not an answer to trust. Name the specific flaw and the input that triggers it, say how you’d catch it (tests, edge cases, reading critically), and how you’d re-prompt or decompose to get it right.
0:00 of about 12 min
Vibe & agentic: describe the solution in plain language (or narrate it) and the coach grades your approach.
Which questions mattered is sealed until you submit. Telling you now would just be handing over the edge cases.
Run or narrate your approach, then ask the coach.