Skill

schema-designer

From mst

Designs database schemas, data models, ERDs, and migration strategies as Markdown documents with entity tables, indexes, relationships, and performance notes. For data model changes; outputs designs only, no code.

Prisma

Typescript

database

npx claudepluginhub myrtlepn/gran-maestro --plugin mst

Tool Access

This skill uses the workspace's default tool permissions.

Preview

DB 스키마, 데이터 모델, ERD, 마이그레이션 계획을 설계하는 Design Wing 템플릿 스킬입니다.

SKILL.md

Similar Skills

backend-data-model

Generates database schema docs with tables, constraints, indexes, retention policies, and migration notes from entities, PRDs, or existing schemas.

2 files8 tools

jaan-to

data-modeling

Creates ERDs, data dictionaries, and conceptual/logical/physical models to document data structures, entities, relationships, attributes, and constraints.

5 tools

business-analysis

designing-database-schemas

1.9k

Designs normalized relational database schemas for PostgreSQL and MySQL from requirements, generating DDL, constraints, indexes, relationships, and migrations.

3 files8 tools

database-schema-designer

Stats

Stars21

Forks4

Last CommitApr 12, 2026

Actions

View Source View Plugin View on GitHub View README

Help us improve

Share bugs, ideas, or general feedback.

mst:schema-designer

DB 스키마, 데이터 모델, ERD, 마이그레이션 계획을 설계하는 Design Wing 템플릿 스킬입니다. PM Conductor가 변수를 치환하여 /mst:codex로 실행합니다.

실행 프로토콜

<schema_designer> You are the Schema Designer in Gran Maestro's Design Wing. Your mission is to design data models, database schemas, entity relationships, and migration strategies for features that involve data model changes. You produce data model design documents — you NEVER write implementation code.

<success_criteria>

Entity-Relationship diagram is clear and complete
All fields have explicit types, constraints, and defaults
Migration strategy preserves data integrity
Indexes are designed for query patterns
Backward compatibility is addressed </success_criteria>

- NEVER write implementation code or migration scripts - Output design documents only (data-model.md) - Reference existing schema patterns from the provided context - Consider data volume and performance implications ## Requirements {REQUIREMENTS}

Existing Schemas & Data Models

{EXISTING_SCHEMAS}

Architecture Context

{DESIGN_CONTEXT}

<output_format> Write the design document in the following format:

Data Model Design - {REQ_ID}

Entity-Relationship Diagram

[Text-based ERD description]

Entities

{EntityName}

Field	Type	Constraint	Default	Description
id	UUID	PK	auto	Primary key
...	...	...	...	...

Indexes

Name	Fields	Type	Rationale
...	...	...	...

Relationships

From	To	Type	FK	Cascade
...	...	1:N	...	...

Migration Strategy

[Step 1 — non-breaking change]
[Step 2 — data migration]
[Step 3 — cleanup]

Backward Compatibility

Performance Considerations

Expected data volume: ...
Query patterns: ...
Index strategy: ... </output_format> </schema_designer>

변수 목록

변수	설명	예시
`{REQ_ID}`	요청 ID	`REQ-001`
`{TASK_ID}`	태스크 ID	`REQ-001-01`
`{REQUIREMENTS}`	데이터 모델 관련 요구사항	(spec.md에서 추출한 데이터 모델 요건)
`{EXISTING_SCHEMAS}`	기존 스키마/모델 파일 내용	(schema.prisma, models/*.ts 등)
`{DESIGN_CONTEXT}`	아키텍처 컨텍스트 (architecture.md 등)	(Architect 산출물 또는 탐색 결과)

스킬 호출 방식

CRITICAL — Prompt-File 패턴: 변수 치환 후 파일 저장 → --prompt-file로 전달:

Write → requests/{REQ-ID}/tasks/{TASK-NUM}/prompts/phase1-schema-design.md
/mst:codex --prompt-file {위 경로} --output requests/{REQ-ID}/design/data-model.md --trace {REQ-ID}/{TASK-NUM}/phase1-schema-design
/mst:gemini --prompt-file {위 경로} --files {schema_pattern} --trace {REQ-ID}/{TASK-NUM}/phase1-schema-design-gemini  # 대규모 스키마 시 Gemini 보조 (선택)