Skill: code-generation

What I do

I provide expertise in using Go's generate tool to automate the creation of boilerplate code. I focus on standard tools like mockgen, stringer, and custom template-based generation to improve maintainability and reduce manual coding.

When to use me

• When adding or updating interface definitions that require new mocks
• When working with enums that need string representation methods
• When implementing repetitive patterns that can be automated via templates

Core principles

1. Automate repetitive tasks: Use generation for code that follows a predictable pattern.
2. Explicit directives: Place //go:generate directives in the files where the source material is defined.
3. Consistency: Ensure generated code follows project style and passes all linting checks.
4. Visibility: Use standard file naming (e.g., _string.go, _mock.go) to distinguish generated files from manual ones.

Patterns & examples

Using stringer for enums:

//go:generate stringer -type=Status
type Status int

const (
    Unknown Status = iota
    Pending
    Active
)

Using mockgen for interfaces:

//go:generate mockgen -destination=mocks/user_repo.go -package=mocks . UserRepository
type UserRepository interface {
    Get(id int) (*User, error)
}

Custom template-based generation: Create a small Go tool that uses the text/template package to generate code from a source definition, then trigger it with //go:generate go run generator.go.

Running generation: Run go generate ./... from the project root to update all generated files.

Anti-patterns to avoid

• ❌ Manual editing: Never edit a generated file. Changes will be overwritten next time go generate runs.
• ❌ Ignoring generated files: Generated code should generally be committed to version control so consumers don't need to install all generation tools.
• ❌ Too much generation: Don't over-engineer solutions. Only generate code when manual maintenance is demonstrably costly or error-prone.

KB Reference

~/vaults/baphled/3. Resources/Knowledge Base/AI Development System/Skills/Code-Quality/Code Generation.md

Related skills

• golang: For idiomatic patterns and template usage
• gomock: Specifics of using the GoMock generation tools
• automation: For integrating generation into CI/CD pipelines