Browse / Developer Tools / Code Validation Workflow

Code Validation Workflow

Validates code changes against format, lint, build, and test checks, and iteratively fixes any failures.

SkillDeveloper ToolsCode QualityOrchestration

The source repository doesn't declare a license. Check its terms before reusing the code.

Key features

  • Iteratively attempts to fix failures and re-validates the code, with a maximum of five attempts.
  • Automatically creates a prioritized TODO list to address failures, from critical compilation errors to minor formatting issues.
  • Spawns subagents to investigate and resolve individual test failures.
  • Runs a comprehensive suite of checks: formatting, linting, building, and testing.
  • Commits fixes with a standardized, co-authored commit message upon successful validation.

Use cases

  • Ensuring code remains stable and passes all checks after refactoring or addressing code review feedback.
  • Verifying code changes pass all quality gates before submitting a pull request.
  • Automatically fixing linting, formatting, and test failures after implementing a new feature.

FAQ

What does the Code Validation Workflow skill do?

This skill automates the process of checking your code for errors. It runs a comprehensive suite of checks—formatting, linting, building, and testing—then creates a prioritized list of any failures and iteratively attempts to fix them.

When should I use this skill?

Use this skill before finalizing a pull request, after implementing new features, or anytime you need to ensure your code is clean, correct, and passes all required quality checks.

How does this skill improve my workflow?

It saves you time by automating the tedious cycle of running checks and fixing common issues. It ensures consistent code quality and helps catch bugs early by running a full validation suite before you commit your changes.

What happens if the skill can't fix an error?

If the skill cannot resolve all validation failures after a maximum of five attempts, it will stop and report the remaining blockers to you, allowing you to manually investigate the more complex issues.