/schema-query

The DB schema says one thing, the queries say another. This skill finds mismatches before they cause production incidents.

What this is for

• Queries referencing columns that don't exist (runtime error)
• Missing indexes on WHERE/JOIN columns (slow at scale)
• Assumed NOT NULL when NULL is allowed (silent wrong results)
• Read-only skill. No database modifications.

PROTECTION RULE - never ~/.claude/

Read-only skill. Guard required if write mode added later.

What You Must Do When Invoked

During analysis, assign a confidence level to each finding: proven (confirmed by evidence), likely (strong signal, needs review), or suspected (weak signal).

If /schema-query -help or /schema-query -h (without further arguments) is invoked: output the ## Usage section unchanged and stop.

Otherwise follow these steps in order, skipping none.

Step 1 - Help check

If invoked with -help or -h, output the ## Usage section unchanged and stop.

Step 2 - Clarify target

Clarify -ProjectDir. Get confirmation.

Step 3 - Scan

& "<SKILL_DIR>/scripts/query-schema-diff.ps1" -ProjectDir "<path>"

Step 4 - Classification

Read each mismatch:

• Critical: runtime error (missing column, type mismatch that causes crash)
• Major: performance at scale (missing index on high-volume query)
• Minor: cosmetic or low-impact (nullable assumption on low-traffic query)
• Info: ORM or framework handles it automatically

Step 5 - Write report

File schema-query-report.md in current working directory:

1. Summary - mismatches by risk level.
2. Mismatch table - critical first. Per mismatch: query file:line, table.column, schema truth, risk, recommendation.
3. Open questions.

Step 6 - Summarize

State report path, highlight mismatches that would cause a runtime error in production.

Usage

/schema-query               # interactive
/schema-query <dir>         # scan project
/schema-query -help