Claude Skill

sample-data

How the sample data is seeded with Bogus behind the Database:SeedSampleData setting, and how tests get their own fixed data. Use when changing seed data, adding a column that needs sample values, or writing tests that depend on known data.

LLM Mart · 0 points · 0 views 0 listing impressions 0 install-command copies
Download leanmeancleanapps-cleanminimalapi-.agents_skills_sample-data-fd0b4df.zip · 1 KB

Install

skills CLI npx skills add https://github.com/LeanMeanCleanApps/CleanMinimalApi/tree/main/.agents/skills/sample-data
Claude Code claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install leanmeancleanapps-cleanminimalapi@llmmart
Git git clone https://github.com/LeanMeanCleanApps/CleanMinimalApi.git

The skills CLI installs just this skill, for any of its supported agents. Claude Code installs the whole leanmeancleanapps/cleanminimalapi collection as a plugin from our marketplace. Git is the plain clone.

Skill manifest

Sample data

The rule

There are two kinds of data, and they never mix.

Where Used by Shape
Sample data src/Infrastructure/Databases/MovieReviews/Extensions/MovieReviewsDbContextExtensions.cs (AddData) Local runs, Docker and the presentation integration tests 15 authors, 50 movies, 150 reviews from Bogus with a fixed seed
Test data tests/Infrastructure.Tests.Integration/Databases/MovieReviews/Extensions/MovieReviewsCollectionData.cs (AddTestData) Repository tests A handful of hand-written rows named "One", "Two", "Three" so assertions can find them

Sample data is only added when Database:SeedSampleData is true (on in appsettings.local.json and appsettings.docker.json, off in appsettings.json; Lambda turns it on only in docker-compose.yml) and the database is empty. A deployed copy never starts with fake data.

Steps

  1. New column on a model? Add a RuleFor to the matching Faker<T> in AddData, and a value to the rows in AddTestData if tests need it.
  2. Movie titles and review text come from the Titles and ReviewTexts lists in AddData. Keep them made up, with no real films, characters or brands, and keep a few titles for common searches such as "space" or "love".
  3. Keep the fixed Seed and ReferenceDate. Dates come from f.Date.Past(refDate: ReferenceDate), never from the clock, so every run gives the same data.
  4. Tests that need a specific row look it up by its known value (Authors.First(a => a.FirstName == "One")), not by position.
  5. Presentation integration tests that need a row with children pick one that has them. Bogus spreads reviews at random, so a few movies and authors have none.

Don't

  • Seed in a repository constructor or on every request. Seeding happens once, in InitialiseInfrastructure().
  • Turn SeedSampleData on in appsettings.json.
  • Use DateTime.Now, Guid.NewGuid() or an unseeded Faker in sample data.
  • Assert exact counts of sample data in presentation tests. They depend on the random spread; repository tests use the test data for exact counts.

Check

dotnet test
Files (cleanminimalapi)
  • SKILL.md 2.4 KB
    ---
    name: sample-data
    description: How the sample data is seeded with Bogus behind the Database:SeedSampleData setting, and how tests get their own fixed data. Use when changing seed data, adding a column that needs sample values, or writing tests that depend on known data.
    metadata:
      scope: shared
    ---
    
    # Sample data
    
    ## The rule
    
    There are two kinds of data, and they never mix.
    
    | | Where | Used by | Shape |
    | --- | --- | --- | --- |
    | Sample data | `src/Infrastructure/Databases/MovieReviews/Extensions/MovieReviewsDbContextExtensions.cs` (`AddData`) | Local runs, Docker and the presentation integration tests | 15 authors, 50 movies, 150 reviews from Bogus with a fixed seed |
    | Test data | `tests/Infrastructure.Tests.Integration/Databases/MovieReviews/Extensions/MovieReviewsCollectionData.cs` (`AddTestData`) | Repository tests | A handful of hand-written rows named "One", "Two", "Three" so assertions can find them |
    
    Sample data is only added when `Database:SeedSampleData` is `true` (on in `appsettings.local.json` and `appsettings.docker.json`, off in `appsettings.json`; Lambda turns it on only in `docker-compose.yml`) and the database is empty. A deployed copy never starts with fake data.
    
    ## Steps
    
    1. New column on a model? Add a `RuleFor` to the matching `Faker<T>` in `AddData`, and a value to the rows in `AddTestData` if tests need it.
    2. Movie titles and review text come from the `Titles` and `ReviewTexts` lists in `AddData`. Keep them made up, with no real films, characters or brands, and keep a few titles for common searches such as "space" or "love".
    3. Keep the fixed `Seed` and `ReferenceDate`. Dates come from `f.Date.Past(refDate: ReferenceDate)`, never from the clock, so every run gives the same data.
    4. Tests that need a specific row look it up by its known value (`Authors.First(a => a.FirstName == "One")`), not by position.
    5. Presentation integration tests that need a row with children pick one that has them. Bogus spreads reviews at random, so a few movies and authors have none.
    
    ## Don't
    
    - Seed in a repository constructor or on every request. Seeding happens once, in `InitialiseInfrastructure()`.
    - Turn `SeedSampleData` on in `appsettings.json`.
    - Use `DateTime.Now`, `Guid.NewGuid()` or an unseeded `Faker` in sample data.
    - Assert exact counts of sample data in presentation tests. They depend on the random spread; repository tests use the test data for exact counts.
    
    ## Check
    
    ``` bash
    dotnet test
    ```
    

Comments (0)

Sign in to join the conversation.

No comments yet.

Reviews (0)

No reviews yet.