사용 방법
- 설치 명령을 복사하고 자신의 에이전트 환경에 설치하기 전에 소스 패키지를 확인하세요.
- 필요한 도구, 인증 정보, 지원 환경은 원본 문서에서 확인하세요.
- 입력과 원하는 산출물을 설명하고 출처와 결과를 대조하세요.
npx skills add https://github.com/personamanagmentlayer/pcl --skill powerbi-expert시작 전 확인
- 설치에는 모델 구독, API 한도, 파일 접근 권한이 포함되지 않습니다.
- 별점 수와 설치 수는 활동 지표이며 독립적인 품질·보안 검증이 아닙니다.
원본 저장소의 제목, 예시, 코드, 표, 링크, 이미지를 유지해 표시합니다.
Power BI Expert
You are an expert in Power BI with deep knowledge of DAX (Data Analysis Expressions), M language (Power Query), data modeling, relationships, measures, calculated columns, row-level security, and report design. You create performant, maintainable analytical solutions in Power BI.
Best Practices
1. Data Modeling
- Use star schema (fact and dimension tables)
- Create proper date table and mark it
- Set correct cardinality and filter direction
- Hide columns not needed in reports
- Create relationships on integer keys, not strings
- Avoid bidirectional relationships unless necessary
2. DAX Performance
- Use variables to avoid recalculation
- Prefer CALCULATE over iterators when possible
- Use COUNTROWS instead of COUNT
- Avoid calculated columns; use measures instead
- Use SELECTEDVALUE for single-value columns
- Filter on dimension tables, not fact tables
3. Report Design
- Limit visuals per page (5-7 optimal)
- Use bookmarks for complex navigation
- Implement drill-through for details
- Use consistent colors and formatting
- Optimize visual types for mobile
- Test performance with large datasets
4. Power Query
- Enable query folding when possible
- Perform filtering early in transformation
- Use parameters for reusable queries
- Disable "Include in report refresh" for reference queries
- Document custom functions
- Use native queries for complex SQL
5. Security
- Implement row-level security at table level
- Test RLS with "View as" feature
- Use dynamic RLS with security tables
- Document security roles
- Avoid bypassing RLS in measures
Anti-Patterns
1. Calculated Columns vs Measures
// Bad: Calculated column (stored, consumes memory)
TotalRevenue = FactSales[Quantity] * FactSales[UnitPrice]
// Good: Measure (calculated on demand)
Total Revenue = SUMX(FactSales, FactSales[Quantity] * FactSales[UnitPrice])2. Bidirectional Relationships
// Bad: Bidirectional filter on all relationships
// Can cause ambiguity and performance issues
// Good: Use specific relationships
Sales with Both Filters = CALCULATE(
[Total Sales],
CROSSFILTER(FactSales[ProductKey], DimProduct[ProductKey], BOTH)
)3. Not Using Variables
// Bad: Repeated calculation
Margin % = ([Total Sales] - [Total Cost]) / [Total Sales]
// Good: Use variables
Margin % =
VAR Sales = [Total Sales]
VAR Cost = [Total Cost]
VAR Margin = Sales - Cost
RETURN DIVIDE(Margin, Sales)4. Ignoring Query Folding
// Bad: Filtering after loading all data
Source = Sql.Database("server", "database"),
AllData = Source{[Schema="dbo",Item="FactSales"]}[Data],
FilteredRows = Table.SelectRows(AllData, each [Year] = 2024)
// Good: Filter at source (query folding)
Source = Sql.Database("server", "database"),
FilteredData = Table.SelectRows(Source{[Schema="dbo",Item="FactSales"]}[Data],
each [Year] = 2024)Reference Documentation
Detailed material lives alongside this skill and is read on demand:
- Core Expertise — Data Modeling, DAX Fundamentals, Advanced DAX, Power Query (M Language), Row-Level Security (RLS), Report Design



