Zurück zur Galerie

Programmierung

SQL Query Review

Review a SQL query for correctness, security, performance, and maintainability.

Modelle
ChatGPT, Claude
Stufe
Fortgeschritten
Version
v0.9

Dieser Prompt ist derzeit auf Englisch verfügbar. Die Oberfläche bleibt in der gewählten Sprache.

01

Was dieser Prompt leistet

Review a SQL query for correctness, security, performance, and maintainability.

02

Geeignet für

Correctness issues, injection risks, index considerations, rewritten query, EXPLAIN plan checklist, and edge cases.

03

Prompt-Variablen

[SQL][SCHEMA][DATABASE][GOAL]
System-Prompt
You are a database reviewer. Treat schema and execution-plan facts as unknown unless supplied.
Benutzer-Prompt
OBJECTIVE:
Review [SQL] against [SCHEMA] on [DATABASE] for the goal [GOAL].

INPUTS:
SQL: [SQL]
SCHEMA: [SCHEMA]
DATABASE: [DATABASE]
GOAL: [GOAL]

OUTPUT CONTRACT:
Correctness issues, injection risks, index considerations, rewritten query, EXPLAIN plan checklist, and edge cases.

Anwendung

Replace every variable with specific, non-sensitive information. Remove irrelevant sections and add one representative example when tone or classification matters.

Warum es funktioniert

The blueprint separates authority, task context, evidence limits, and the output contract so requirements can be checked independently.

Modellkompatibilität

ChatGPTuntestedEditorial compatibility hypothesis; cross-model test record pending.

ClaudeuntestedEditorial compatibility hypothesis; cross-model test record pending.

05

Beispieleingabe

SQL: [SQL]
SCHEMA: [SCHEMA]
DATABASE: [DATABASE]
GOAL: [GOAL]

06

Beispielausgabe

A structured deliverable following the listed output contract, with assumptions and missing evidence clearly marked.

Anpassen

Replace every variable with specific, non-sensitive information. Remove irrelevant sections and add one representative example when tone or classification matters.

Häufige Fehler

Do not paste secrets or personal data. Avoid contradictory constraints, unsupported facts, and using generated output without review.

Testnotizen

Editorial blueprint awaiting recorded cross-model evaluation. It is visible for review but is not labeled as tested.

Why is this blueprint not marked as tested?

It has passed editorial structure review but does not yet have a published cross-model test record.