Skill UI

This skill performs security-focused code reviews for pull requests, commits, and diffs. It prioritizes high-risk areas like authentication, cryptography, and external calls. The process is evidence-based, requiring specific line numbers and attack scenarios. It adapts to codebase size and always generates a comprehensive markdown report.

This skill guides users on setting up automated code quality gates directly at the Git level. It covers best practices using frameworks like Husky, lint-staged, and pre-commit. Users can enforce linting, formatting, type-checking, and standardized commit message conventions (e.g., Conventional Commits) before code is committed or pushed, significantly improving code hygiene and development workflow consistency.

This skill analyzes the comprehensive commit history between two branches to generate a concise, objective, and highly structured Pull Request (PR) description. It automatically groups changes by domain, summarizing the purpose, technical notes, and user impact, ensuring consistency and minimizing manual documentation effort.

Automate software version management and changelog generation based on Semantic Versioning principles and Conventional Commits. Supports tools like semantic-release and standard-version for CI/CD integration, ensuring consistent release workflows and version precedence.

Rewrites commit messages to sound like a careful human engineer wrote them. Strips AI/marketing slop (e.g., 'comprehensive solution', 'robust implementation', 'leverage') while keeping Conventional Commits format. Ensures direct, specific, and imperative subject lines with optional context-rich body. Ideal for developers who want clean, professional git history without AI-generated jargon.

Generates conventional-commit messages from staged changes: a type prefix plus an English imperative subject (≤50 chars) and an optional why-focused body. Use when asked to write, generate, or polish a git commit message.