mcpField
C
Code Formatting

Code Formatting

github.com/openshiftTesting & quality54882 installsUpdated 1d ago
Repository

About Code Formatting

Code Formatting is a skill for code quality and testing conventions in Go projects. It specifies formatting and verification commands, test naming, and test placement.

Key features

  1. 01Fix lint issues with make lint-fix
  2. 02Verify code and tests with make verify
  3. 03Specify Go test case naming
  4. 04Require unit tests for changed functions
  5. 05Explain where tests belong

Use cases

  1. 01Run quality checks before committing Go code
  2. 02Write and place unit tests for new functions
  3. 03Check spelling in Markdown

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/openshift/hypershift --skill code-formatting

After installing, type/code-formattingin Claude Code to confirm it loaded.

Details

From the project README

Code Quality, Formatting and Conventions

Code Formatting and Linting

  • Use make lint-fix after writing Go code to automatically fix most linting issues
  • Run make verify to verify both linting and tests pass before committing
  • For markdown files, use make verify-codespell to catch spelling errors

Test Conventions - MANDATORY

Test naming

NON-NEGOTIABLE RULE: Every Go test name MUST follow this exact format:

name: "When <condition>, it should <expected behavior>"

Examples:

name: "When NodePool has valid image ID, it should create Azure machine template"
name: "When subnet ID is invalid, it should return error"
name: "When encryption is enabled, it should configure disk encryption set"

NEVER use generic names like:

  • ❌ "nominal case without managed identity"
  • ❌ "basic Azure machine template with ImageID"
  • ❌ "error case - invalid subnet ID"

ALWAYS use "When...it should..." format:

  • ✅ "When NodePool has no managed identity, it should create template without identity"
  • ✅ "When NodePool has ImageID, it should create basic Azure machine template"
  • ✅ "When subnet ID is invalid, it should return error with subnet parse message"
Unit test function naming

Unit test functions MUST be named after the function they test, using the standard Go Test<FunctionName> convention:

// Testing function ReconcileNodePool
func TestReconcileNodePool(t *testing.T) { ... }

// Testing function buildAzureMachineTemplate
func TestBuildAzureMachineTemplate(t *testing.T) { ... }

NEVER use generic or disconnected test function names like:

  • ❌ TestNodePoolFeatures
  • ❌ TestAzureIntegration
  • ❌ TestVariousCases

Each test function should map 1:1 to the function under test. If a function is complex enough to warrant many test cases, use table-driven tests within a single Test<FunctionName> function with the "When...it should..." naming for each case.

Unit test placement
  • Always include unit tests when creating new functions or modifying existing ones
  • Place unit tests next to the code they test — in the same package directory, in a _test.go file
  • Do NOT place new tests in test/integration/ — this is a legacy directory. New unit tests go alongside the code under test. New integration-style tests that require a cluster should use test/e2e/ or test/envtest/

Quick Checklist

Before committing:

  • Ran make lint-fix on Go code
  • Ran make verify (passes)
  • Tests use "When...it should..." naming format
  • New functions have unit tests
  • Markdown checked with make verify-codespell

Comments

Markdown supported