Data Fixture Auditor

Test fixtures silently rot: happy-path only, stale shapes, zero error-scenario coverage. This skill scans fixture files, parses entities, measures field coverage, and generates a quality report.

PROTECTION RULE - never ~/.claude/

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

Usage

/fixture-audit <project-directory>

Loade das Skill, dann:

/fixture-audit $ProjectDir\tests\fixtures\smoke\src

What You Must Do When Invoked

Step 1

Run /fixture-audit <project-directory> against the target project.

Step 2

Review the fixture quality report for coverage gaps identified by the scan.

Step 3

Document recommendations to improve fixture coverage.

Pipeline

1. Collect - fixture-scan.ps1 scannt nach Fixture-Dateien (JSON, YAML, SQL INSERT, Factory-Definitions) anhand Namenskonvention (*fixture*, *seed*, *test-data*).
2. Parse - Extrahiert Entities und Feldwerte aus jeder Datei.
3. Measure - Pro Entity: Cardinality, Konstante-Felder, Null-Ratio, Coverage.
4. Validate - Vergleich gegen Companion-Schema (DDL oder JSON Schema).
5. Report - fixture-quality-report.md mit LLM-bewerteter Coverage-Analyse.

Quality Dimensions

Dimension	Bedeutung
Field coverage	Welche Columns sind populated vs. immer null vs. fehlen
Cardinality	Verschiedene Werte pro Feld (niedrig = Testlücke)
Constant fields	Gleicher Wert in jedem Record
Schema drift	Felder im Schema aber nie in Fixtures

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