devskill 2.0.6 → 2.0.8

package/skills/powerbi-modeling/SKILL.md

@@ -0,0 +1,153 @@
+---
+name: powerbi-modeling
+description: 'Power BI semantic modeling assistant for building optimized data models. Use when working with Power BI semantic models, creating measures, designing star schemas, configuring relationships, implementing RLS, or optimizing model performance. Triggers on queries about DAX calculations, table relationships, dimension/fact table design, naming conventions, model documentation, cardinality, cross-filter direction, calculation groups, and data model best practices. Always connects to the active model first using power-bi-modeling MCP tools to understand the data structure before providing guidance.'
+---
+
+# Power BI Semantic Modeling
+
+Guide users in building optimized, well-documented Power BI semantic models following Microsoft best practices.
+
+## When to Use This Skill
+
+Use this skill when users ask about:
+- Creating or optimizing Power BI semantic models
+- Designing star schemas (dimension/fact tables)
+- Writing DAX measures or calculated columns
+- Configuring table relationships (cardinality, cross-filter)
+- Implementing row-level security (RLS)
+- Naming conventions for tables, columns, measures
+- Adding descriptions and documentation to models
+- Performance tuning and optimization
+- Calculation groups and field parameters
+- Model validation and best practice checks
+
+**Trigger phrases:** "create a measure", "add relationship", "star schema", "optimize model", "DAX formula", "RLS", "naming convention", "model documentation", "cardinality", "cross-filter"
+
+## Prerequisites
+
+### Required Tools
+- **Power BI Modeling MCP Server**: Required for connecting to and modifying semantic models
+  - Enables: connection_operations, table_operations, measure_operations, relationship_operations, etc.
+  - Must be configured and running to interact with models
+
+### Optional Dependencies
+- **Microsoft Learn MCP Server**: Recommended for researching latest best practices
+  - Enables: microsoft_docs_search, microsoft_docs_fetch
+  - Use for complex scenarios, new features, and official documentation
+
+## Workflow
+
+### 1. Connect and Analyze First
+
+Before providing any modeling guidance, always examine the current model state:
+
+```
+1. List connections: connection_operations(operation: "ListConnections")
+2. If no connection, check for local instances: connection_operations(operation: "ListLocalInstances")
+3. Connect to the model (Desktop or Fabric)
+4. Get model overview: model_operations(operation: "Get")
+5. List tables: table_operations(operation: "List")
+6. List relationships: relationship_operations(operation: "List")
+7. List measures: measure_operations(operation: "List")
+```
+
+### 2. Evaluate Model Health
+
+After connecting, assess the model against best practices:
+
+- **Star Schema**: Are tables properly classified as dimension or fact?
+- **Relationships**: Correct cardinality? Minimal bidirectional filters?
+- **Naming**: Human-readable, consistent naming conventions?
+- **Documentation**: Do tables, columns, measures have descriptions?
+- **Measures**: Explicit measures for key calculations?
+- **Hidden Fields**: Are technical columns hidden from report view?
+
+### 3. Provide Targeted Guidance
+
+Based on analysis, guide improvements using references:
+- Star schema design: See [STAR-SCHEMA.md](references/STAR-SCHEMA.md)
+- Relationship configuration: See [RELATIONSHIPS.md](references/RELATIONSHIPS.md)
+- DAX measures and naming: See [MEASURES-DAX.md](references/MEASURES-DAX.md)
+- Performance optimization: See [PERFORMANCE.md](references/PERFORMANCE.md)
+- Row-level security: See [RLS.md](references/RLS.md)
+
+## Quick Reference: Model Quality Checklist
+
+| Area | Best Practice |
+|------|--------------|
+| Tables | Clear dimension vs fact classification |
+| Naming | Human-readable: `Customer Name` not `CUST_NM` |
+| Descriptions | All tables, columns, measures documented |
+| Measures | Explicit DAX measures for business metrics |
+| Relationships | One-to-many from dimension to fact |
+| Cross-filter | Single direction unless specifically needed |
+| Hidden fields | Hide technical keys, IDs from report view |
+| Date table | Dedicated marked date table |
+
+## MCP Tools Reference
+
+Use these Power BI Modeling MCP operations:
+
+| Operation Category | Key Operations |
+|-------------------|----------------|
+| `connection_operations` | Connect, ListConnections, ListLocalInstances, ConnectFabric |
+| `model_operations` | Get, GetStats, ExportTMDL |
+| `table_operations` | List, Get, Create, Update, GetSchema |
+| `column_operations` | List, Get, Create, Update (descriptions, hidden, format) |
+| `measure_operations` | List, Get, Create, Update, Move |
+| `relationship_operations` | List, Get, Create, Update, Activate, Deactivate |
+| `dax_query_operations` | Execute, Validate |
+| `calculation_group_operations` | List, Create, Update |
+| `security_role_operations` | List, Create, Update, GetEffectivePermissions |
+
+## Common Tasks
+
+### Add Measure with Description
+```
+measure_operations(
+  operation: "Create",
+  definitions: [{
+    name: "Total Sales",
+    tableName: "Sales",
+    expression: "SUM(Sales[Amount])",
+    formatString: "$#,##0",
+    description: "Sum of all sales amounts"
+  }]
+)
+```
+
+### Update Column Description
+```
+column_operations(
+  operation: "Update",
+  definitions: [{
+    tableName: "Customer",
+    name: "CustomerKey",
+    description: "Unique identifier for customer dimension",
+    isHidden: true
+  }]
+)
+```
+
+### Create Relationship
+```
+relationship_operations(
+  operation: "Create",
+  definitions: [{
+    fromTable: "Sales",
+    fromColumn: "CustomerKey",
+    toTable: "Customer",
+    toColumn: "CustomerKey",
+    crossFilteringBehavior: "OneDirection"
+  }]
+)
+```
+
+## When to Use Microsoft Learn MCP
+
+Research current best practices using `microsoft_docs_search` for:
+- Latest DAX function documentation
+- New Power BI features and capabilities
+- Complex modeling scenarios (SCD Type 2, many-to-many)
+- Performance optimization techniques
+- Security implementation patterns

package/skills/powerbi-modeling/SYNC.md

@@ -0,0 +1,5 @@
+# Sync Info
+
+- **Source:** `vendor/awesome-copilot/skills/powerbi-modeling`
+- **Git SHA:** `23fb7006c132a0bba75b0e5af7a35cab11272812`
+- **Synced:** 2026-03-24

package/skills/powerbi-modeling/references/MEASURES-DAX.md

@@ -0,0 +1,195 @@
+# DAX Measures and Naming Conventions
+
+## Naming Conventions
+
+### General Rules
+- Use human-readable names (spaces allowed)
+- Be descriptive: `Total Sales Amount` not `TSA`
+- Avoid abbreviations unless universally understood
+- Use consistent capitalization (Title Case recommended)
+- Avoid special characters except spaces
+
+### Table Naming
+| Type | Convention | Example |
+|------|------------|---------|
+| Dimension | Singular noun | Customer, Product, Date |
+| Fact | Business process | Sales, Orders, Inventory |
+| Bridge | Combined names | CustomerAccount, ProductCategory |
+| Measure Table | Underscore prefix | _Measures, _KPIs |
+
+### Column Naming
+| Type | Convention | Example |
+|------|------------|---------|
+| Keys | Suffix with "Key" or "ID" | CustomerKey, ProductID |
+| Dates | Suffix with "Date" | OrderDate, ShipDate |
+| Amounts | Descriptive with unit hint | SalesAmount, QuantitySold |
+| Flags | Prefix with "Is" or "Has" | IsActive, HasDiscount |
+
+### Measure Naming
+| Type | Convention | Example |
+|------|------------|---------|
+| Aggregations | Verb + Noun | Total Sales, Count of Orders |
+| Ratios | X per Y or X Rate | Sales per Customer, Conversion Rate |
+| Time Intelligence | Period + Metric | YTD Sales, PY Total Sales |
+| Comparisons | Metric + vs + Baseline | Sales vs Budget, Growth vs PY |
+
+## Explicit vs Implicit Measures
+
+### Always Create Explicit Measures For:
+1. Key business metrics users will query
+2. Complex calculations with filter manipulation
+3. Measures used in MDX (Excel PivotTables)
+4. Controlled aggregation (prevent sum of averages)
+
+### Implicit Measures (Column Aggregations)
+- Acceptable for simple exploration
+- Set correct SummarizeBy property:
+  - Amounts: Sum
+  - Keys/IDs: None (Do Not Summarize)
+  - Rates/Prices: None or Average
+
+## Measure Patterns
+
+### Basic Aggregations
+```dax
+Total Sales = SUM(Sales[SalesAmount])
+Order Count = COUNTROWS(Sales)
+Average Order Value = DIVIDE([Total Sales], [Order Count])
+Distinct Customers = DISTINCTCOUNT(Sales[CustomerKey])
+```
+
+### Time Intelligence (Requires Date Table)
+```dax
+YTD Sales = TOTALYTD([Total Sales], 'Date'[Date])
+MTD Sales = TOTALMTD([Total Sales], 'Date'[Date])
+PY Sales = CALCULATE([Total Sales], SAMEPERIODLASTYEAR('Date'[Date]))
+YoY Growth = DIVIDE([Total Sales] - [PY Sales], [PY Sales])
+```
+
+### Percentage Calculations
+```dax
+Sales % of Total = 
+DIVIDE(
+    [Total Sales],
+    CALCULATE([Total Sales], REMOVEFILTERS(Product))
+)
+
+Margin % = DIVIDE([Gross Profit], [Total Sales])
+```
+
+### Running Totals
+```dax
+Running Total = 
+CALCULATE(
+    [Total Sales],
+    FILTER(
+        ALL('Date'),
+        'Date'[Date] <= MAX('Date'[Date])
+    )
+)
+```
+
+## Column References
+
+### Best Practice: Always Qualify Column Names
+```dax
+// GOOD - Fully qualified
+Sales Amount = SUM(Sales[SalesAmount])
+
+// BAD - Unqualified (can cause ambiguity)
+Sales Amount = SUM([SalesAmount])
+```
+
+### Measure References: Never Qualify
+```dax
+// GOOD - Unqualified measure
+YTD Sales = TOTALYTD([Total Sales], 'Date'[Date])
+
+// BAD - Qualified measure (breaks if home table changes)
+YTD Sales = TOTALYTD(Sales[Total Sales], 'Date'[Date])
+```
+
+## Documentation
+
+### Measure Descriptions
+Always add descriptions explaining:
+- What the measure calculates
+- Business context/usage
+- Any important assumptions
+
+```
+measure_operations(
+  operation: "Update",
+  definitions: [{
+    name: "Total Sales",
+    tableName: "Sales",
+    description: "Sum of all completed sales transactions. Excludes returns and cancelled orders."
+  }]
+)
+```
+
+### Format Strings
+| Data Type | Format String | Example Output |
+|-----------|---------------|----------------|
+| Currency | $#,##0.00 | $1,234.56 |
+| Percentage | 0.0% | 12.3% |
+| Whole Number | #,##0 | 1,234 |
+| Decimal | #,##0.00 | 1,234.56 |
+
+## Display Folders
+
+Organize measures into logical groups:
+```
+measure_operations(
+  operation: "Update",
+  definitions: [{
+    name: "YTD Sales",
+    tableName: "_Measures",
+    displayFolder: "Time Intelligence\\Year"
+  }]
+)
+```
+
+Common folder structure:
+```
+_Measures
+├── Sales
+│   ├── Total Sales
+│   └── Average Sale
+├── Time Intelligence
+│   ├── Year
+│   │   ├── YTD Sales
+│   │   └── PY Sales
+│   └── Month
+│       └── MTD Sales
+└── Ratios
+    ├── Margin %
+    └── Conversion Rate
+```
+
+## Variables for Performance
+
+Use variables to:
+- Avoid recalculating the same expression
+- Improve readability
+- Enable debugging
+
+```dax
+Gross Margin % = 
+VAR TotalSales = [Total Sales]
+VAR TotalCost = [Total Cost]
+VAR GrossProfit = TotalSales - TotalCost
+RETURN
+    DIVIDE(GrossProfit, TotalSales)
+```
+
+## Validation Checklist
+
+- [ ] All key business metrics have explicit measures
+- [ ] Measures have clear, descriptive names
+- [ ] Measures have descriptions
+- [ ] Appropriate format strings applied
+- [ ] Display folders organize related measures
+- [ ] Column references are fully qualified
+- [ ] Measure references are not qualified
+- [ ] Variables used for complex calculations

package/skills/powerbi-modeling/references/PERFORMANCE.md

@@ -0,0 +1,215 @@
+# Performance Optimization for Power BI Models
+
+## Data Reduction Techniques
+
+### 1. Remove Unnecessary Columns
+- Only import columns needed for reporting
+- Remove audit columns (CreatedBy, ModifiedDate) unless required
+- Remove duplicate/redundant columns
+
+```
+column_operations(operation: "List", filter: { tableNames: ["Sales"] })
+// Review and remove unneeded columns
+```
+
+### 2. Remove Unnecessary Rows
+- Filter historical data to relevant period
+- Exclude cancelled/void transactions if not needed
+- Apply filters in Power Query (not in DAX)
+
+### 3. Reduce Cardinality
+High cardinality (many unique values) impacts:
+- Model size
+- Refresh time
+- Query performance
+
+**Solutions:**
+| Column Type | Reduction Technique |
+|-------------|---------------------|
+| DateTime | Split into Date and Time columns |
+| Decimal precision | Round to needed precision |
+| Text with patterns | Extract common prefix/suffix |
+| High-precision IDs | Use surrogate integer keys |
+
+### 4. Optimize Data Types
+| From | To | Benefit |
+|------|-----|---------|
+| DateTime | Date (if time not needed) | 8 bytes to 4 bytes |
+| Decimal | Fixed Decimal | Better compression |
+| Text with numbers | Whole Number | Much better compression |
+| Long text | Shorter text | Reduces storage |
+
+### 5. Group and Summarize
+Pre-aggregate data when detail not needed:
+- Daily instead of transactional
+- Monthly instead of daily
+- Consider aggregation tables for DirectQuery
+
+## Column Optimization
+
+### Prefer Power Query Columns Over Calculated Columns
+| Approach | When to Use |
+|----------|-------------|
+| Power Query (M) | Can be computed at source, static values |
+| Calculated Column (DAX) | Needs model relationships, dynamic logic |
+
+Power Query columns:
+- Load faster
+- Compress better
+- Use less memory
+
+### Avoid Calculated Columns on Relationship Keys
+DAX calculated columns in relationships:
+- Cannot use indexes
+- Generate complex SQL for DirectQuery
+- Hurt performance significantly
+
+**Use COMBINEVALUES for multi-column relationships:**
+```dax
+// If you must use calculated column for composite key
+CompositeKey = COMBINEVALUES(",", [Country], [City])
+```
+
+### Set Appropriate Summarization
+Prevent accidental aggregation of non-additive columns:
+```
+column_operations(
+  operation: "Update",
+  definitions: [{
+    tableName: "Product",
+    name: "UnitPrice",
+    summarizeBy: "None"
+  }]
+)
+```
+
+## Relationship Optimization
+
+### 1. Minimize Bidirectional Relationships
+Each bidirectional relationship:
+- Increases query complexity
+- Can create ambiguous paths
+- Reduces performance
+
+### 2. Avoid Many-to-Many When Possible
+Many-to-many relationships:
+- Generate more complex queries
+- Require more memory
+- Can produce unexpected results
+
+### 3. Reduce Relationship Cardinality
+Keep relationship columns low cardinality:
+- Use integer keys over text
+- Consider higher-grain relationships
+
+## DAX Optimization
+
+### 1. Use Variables
+```dax
+// GOOD - Calculate once, use twice
+Sales Growth = 
+VAR CurrentSales = [Total Sales]
+VAR PriorSales = [PY Sales]
+RETURN DIVIDE(CurrentSales - PriorSales, PriorSales)
+
+// BAD - Recalculates [Total Sales] and [PY Sales]
+Sales Growth = 
+DIVIDE([Total Sales] - [PY Sales], [PY Sales])
+```
+
+### 2. Avoid FILTER with Entire Tables
+```dax
+// BAD - Iterates entire table
+Sales High Value = 
+CALCULATE([Total Sales], FILTER(Sales, Sales[Amount] > 1000))
+
+// GOOD - Uses column reference
+Sales High Value = 
+CALCULATE([Total Sales], Sales[Amount] > 1000)
+```
+
+### 3. Use KEEPFILTERS Appropriately
+```dax
+// Respects existing filters
+Sales with Filter = 
+CALCULATE([Total Sales], KEEPFILTERS(Product[Category] = "Bikes"))
+```
+
+### 4. Prefer DIVIDE Over Division Operator
+```dax
+// GOOD - Handles divide by zero
+Margin % = DIVIDE([Profit], [Sales])
+
+// BAD - Errors on zero
+Margin % = [Profit] / [Sales]
+```
+
+## DirectQuery Optimization
+
+### 1. Minimize Columns and Tables
+DirectQuery models:
+- Query source for every visual
+- Performance depends on source
+- Minimize data retrieved
+
+### 2. Avoid Complex Power Query Transformations
+- Transforms become subqueries
+- Native queries are faster
+- Materialize at source when possible
+
+### 3. Keep Measures Simple Initially
+Complex DAX generates complex SQL:
+- Start with basic aggregations
+- Add complexity gradually
+- Monitor query performance
+
+### 4. Disable Auto Date/Time
+For DirectQuery models, disable auto date/time:
+- Creates hidden calculated tables
+- Increases model complexity
+- Use explicit date table instead
+
+## Aggregations
+
+### User-Defined Aggregations
+Pre-aggregate fact tables for:
+- Very large models (billions of rows)
+- Hybrid DirectQuery/Import
+- Common query patterns
+
+```
+table_operations(
+  operation: "Create",
+  definitions: [{
+    name: "SalesAgg",
+    mode: "Import",
+    mExpression: "..."
+  }]
+)
+```
+
+## Performance Testing
+
+### Use Performance Analyzer
+1. Enable in Power BI Desktop
+2. Start recording
+3. Interact with visuals
+4. Review DAX query times
+
+### Monitor with DAX Studio
+External tool for:
+- Query timing
+- Server timings
+- Query plans
+
+## Validation Checklist
+
+- [ ] Unnecessary columns removed
+- [ ] Appropriate data types used
+- [ ] High-cardinality columns addressed
+- [ ] Bidirectional relationships minimized
+- [ ] DAX uses variables for repeated expressions
+- [ ] No FILTER on entire tables
+- [ ] DIVIDE used instead of division operator
+- [ ] Auto date/time disabled for DirectQuery
+- [ ] Performance tested with representative data

package/skills/powerbi-modeling/references/RELATIONSHIPS.md

@@ -0,0 +1,147 @@
+# Relationships in Power BI
+
+## Relationship Properties
+
+### Cardinality
+| Type | Use Case | Notes |
+|------|----------|-------|
+| One-to-Many (*:1) | Dimension to Fact | Most common, preferred |
+| Many-to-One (1:*) | Fact to Dimension | Same as above, direction reversed |
+| One-to-One (1:1) | Dimension extensions | Use sparingly |
+| Many-to-Many (*:*) | Bridge tables, complex scenarios | Requires careful design |
+
+### Cross-Filter Direction
+| Setting | Behavior | When to Use |
+|---------|----------|-------------|
+| Single | Filters flow from "one" to "many" | Default, best performance |
+| Both | Filters flow in both directions | Only when necessary |
+
+## Best Practices
+
+### 1. Prefer One-to-Many Relationships
+```
+Customer (1) --> (*) Sales
+Product  (1) --> (*) Sales
+Date     (1) --> (*) Sales
+```
+
+### 2. Use Single-Direction Cross-Filtering
+Bidirectional filtering:
+- Impacts performance negatively
+- Can create ambiguous filter paths
+- May produce unexpected results
+
+**Only use bidirectional when:**
+- Dimension-to-dimension analysis through fact table
+- Specific RLS requirements
+
+**Better alternative:** Use CROSSFILTER in DAX measures:
+```dax
+Countries Sold = 
+CALCULATE(
+    DISTINCTCOUNT(Customer[Country]),
+    CROSSFILTER(Customer[CustomerKey], Sales[CustomerKey], BOTH)
+)
+```
+
+### 3. One Active Path Between Tables
+- Only one active relationship between any two tables
+- Use USERELATIONSHIP for role-playing dimensions:
+
+```dax
+Sales by Ship Date = 
+CALCULATE(
+    [Total Sales],
+    USERELATIONSHIP(Sales[ShipDate], Date[Date])
+)
+```
+
+### 4. Avoid Ambiguous Paths
+Circular references cause errors. Solutions:
+- Deactivate one relationship
+- Restructure model
+- Use USERELATIONSHIP in measures
+
+## Relationship Patterns
+
+### Standard Star Schema
+```
+     [Date]
+       |
+[Product]--[Sales]--[Customer]
+       |
+   [Store]
+```
+
+### Role-Playing Dimension
+```
+[Date] --(active)-- [Sales.OrderDate]
+   |
+   +--(inactive)-- [Sales.ShipDate]
+```
+
+### Bridge Table (Many-to-Many)
+```
+[Customer]--(*)--[CustomerAccount]--(*)--[Account]
+```
+
+### Factless Fact Table
+```
+[Product]--[ProductPromotion]--[Promotion]
+```
+Used to capture relationships without measures.
+
+## Creating Relationships via MCP
+
+### List Current Relationships
+```
+relationship_operations(operation: "List")
+```
+
+### Create New Relationship
+```
+relationship_operations(
+  operation: "Create",
+  definitions: [{
+    fromTable: "Sales",
+    fromColumn: "ProductKey",
+    toTable: "Product", 
+    toColumn: "ProductKey",
+    crossFilteringBehavior: "OneDirection",
+    isActive: true
+  }]
+)
+```
+
+### Deactivate Relationship
+```
+relationship_operations(
+  operation: "Deactivate",
+  references: [{ name: "relationship-guid-here" }]
+)
+```
+
+## Troubleshooting
+
+### "Ambiguous Path" Error
+Multiple active paths exist between tables.
+- Check for: Multiple fact tables sharing dimensions
+- Solution: Deactivate redundant relationships
+
+### Bidirectional Not Allowed
+Circular reference would be created.
+- Solution: Restructure or use DAX CROSSFILTER
+
+### Relationship Not Detected
+Columns may have different data types.
+- Ensure both columns have identical types
+- Check for trailing spaces in text keys
+
+## Validation Checklist
+
+- [ ] All relationships are one-to-many where possible
+- [ ] Cross-filter is single direction by default
+- [ ] Only one active path between any two tables
+- [ ] Role-playing dimensions use inactive relationships
+- [ ] No circular reference paths
+- [ ] Key columns have matching data types