메뉴
Analyze change impact and combination regression for existing C# batches backed by SQL Server stored procedures.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Structure brownfield change work across requirements, planning, design, manufacturing, and testing by re-organizing upstream items into phase results, decisions, unknowns, evidence, gates, and handoffs. Use when an existing codebase or system must be changed without treating its current implementation as the complete specification.
Run isolated calibration and drift evaluations for PyPI-distributed XRefKit Skills. Use when installed Skill packages expose evaluation manifests and the evaluator must run fixtures repeatedly without revealing expected answers, calibration rules, or held-out cases to the evaluated Skill.
Analyze change impact and combination regression for existing C# batches backed by SQL Server stored procedures. Use when a business requirement changes a batch whose rules, validation, calculations, updates, or combination validity may be split between C# and SQL Server, especially when the candidate space is large and no formal test dataset exists.
Create repository-ready marketing explainer videos from a message flow using staged slide reveals, dialogue-style narration, HTML/CSS slide rendering, TTS audio, licensing credits, preview pages, and README/video placement. Use when Codex is asked to make or revise a short explanatory video, overview video, narrated slide video, Japanese/English marketing video, or README-linked product video.
| name | batch-impact-regression |
| description | Analyze change impact and combination regression for existing C# batches backed by SQL Server stored procedures. |
Use this Skill when a requirement changes an existing C# batch and SQL Server stored-procedure execution path. Inspect C# and SQL as one execution path, trace parameters and database dependencies, and keep business constraints separate from source conditions.
Follow the 15-step workflow in references/workflow.md. Use the deterministic
tools in scripts/ for combination generation, constraint application,
source decision-table extraction, result normalization, comparison, and
regression-set selection. Never send the full candidate space to the model.
python scripts/code_tables.py C:\path\to\source -o tables.json
python scripts/batch_regression.py report config.json old.json new.json -o report.json
The real batch and database require an isolated adapter. Do not execute against production. Treat dynamic SQL, unresolved dynamic procedure names, missing business evidence, and unexplained differences as unknowns requiring human judgment. The baseline is observed current behavior, not proof of business correctness.