How does this skill differ from running a standard linter?

Unlike basic linters, this skill performs a 'before vs. after' comparison to detect new warnings and identifies complex SonarCloud-specific smells like nested template literals and improper React key usage.

Does this skill work with monorepos?

Yes, it is designed to recognize and validate multiple subsystems including React packages, CLI tools, MCP servers, and documentation folders.

Why does the skill use git stash during execution?

It stashes changes to establish a baseline for linting and testing, allowing it to accurately report if your new changes introduced any regressions compared to the previous state.

Can it detect build output errors?

Yes, it specifically checks generated files for common issues such as '[object Object]' strings caused by template literal bugs and verifies the presence of proper ESM/CJS exports.

Pre-PR Deep Review

Name: Pre-PR Deep Review
Author: mkatogui

bymkatogui

•

部署与 DevOps

Performs comprehensive code validation and CI simulation to catch errors, lint regressions, and SonarCloud smells before PR submission.

Pre-PR Deep Review is a robust validation skill designed to act as a local CI pipeline. It executes a multi-step verification process that includes TypeScript compilation checks, automated linting baseline comparisons, and deep diff reviews to identify specific SonarCloud code smells like nested template literals or improper React keys. By verifying build outputs and running targeted test suites across multiple subsystems, it ensures your code meets high-quality standards and will pass remote repository quality gates without revision.

主要功能

01SonarCloud-specific pattern matching for rules often missed by standard linters

02Detailed readiness reporting with clear blocker and warning classifications

03Automated lint baseline comparison to prevent regression in warning counts

04Validation of build outputs for serialization errors like '[object Object]'

051 GitHub stars

06Sequential execution of TypeScript, lint, and test suites across multiple subsystems

使用场景

01Validating complex changes across monorepo subsystems before opening a pull request

02Debugging CI pipeline failures locally by simulating strict SonarCloud and Biome quality gates

03Ensuring design system token builds generate valid ESM and CJS exports

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add mkatogui/universal-design-system pre-pr-review

For use in Claude.ai and ChatGPT

主要功能

01SonarCloud-specific pattern matching for rules often missed by standard linters

02Detailed readiness reporting with clear blocker and warning classifications

03Automated lint baseline comparison to prevent regression in warning counts

04Validation of build outputs for serialization errors like '[object Object]'

051 GitHub stars

06Sequential execution of TypeScript, lint, and test suites across multiple subsystems

使用场景

01Validating complex changes across monorepo subsystems before opening a pull request

02Debugging CI pipeline failures locally by simulating strict SonarCloud and Biome quality gates

03Ensuring design system token builds generate valid ESM and CJS exports

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add mkatogui/universal-design-system pre-pr-review

For use in Claude.ai and ChatGPT