Найдите подходящую возможность для вашего агента.
Uses ast-grep to find patterns and analyze code for refactoring. Use when planning refactors, finding deprecated API usage, migrating code, or analyzing code quality before changes. For simple code search, use searching-code instead.
Analyzes code complexity metrics and identifies areas that are too complex for developers to easily understand
Validate and enforce markdown document structure including frontmatter positioning, heading hierarchy, and content organization. Use when creating or validating markdown files.
Conduct a detailed technical review of recent changes or a specific feature. Use when user says "review PR", "code review", or "review [scope]". Analyzes for logic errors, bugs, style issues, and performance. Fixes critical/high issues automatically.
Checks for breaking changes in API serializers before production release. Use when reviewing PRs to production branch.
Validate Swift Packages against ARC Labs Studio standards defined in ARCKnowledge. Use when checking package compliance, auditing structure, reviewing before release, or verifying ARCDevTools integration. Triggers on requests like "validate package", "check ARC standards", "is this package ready", "audit ARCPackage", "package compliance check", "review package structure".
Validate and fix folder structure for Output SDK workflows. Use to ensure migrated workflows follow the correct structure conventions.
This skill should be used when detecting and fixing color contrast issues in the codebase. It finds dark text on dark backgrounds and light text on light backgrounds, then provides fixes to ensure proper readability. Uses an automated Python script for detection and provides Tailwind-specific fix recommendations.
Trigger when the user requests a review of frontend files (e.g., `.tsx`, `.ts`, `.js`). Support both pending-change reviews and focused file reviews.
Conducts comprehensive backend code reviews including API design (REST/GraphQL/gRPC), database patterns, authentication/authorization, caching strategies, message queues, microservices architecture, security vulnerabilities, and performance optimization for Node.js, Python, Java, Go, and C#. Produces detailed review reports with specific issues, severity ratings, and actionable recommendations. Use when reviewing server-side code, analyzing API implementations, checking database queries, validating authentication flows, assessing microservices architecture, or when users mention "review backend code", "check API design", "analyze server code", "validate database patterns", "security audit", "performance review", or "backend code quality".
Enforce code complexity limits and quality standards in TapScore backend. Use when writing or reviewing backend code to ensure method size limits, nesting depth, naming conventions, and use of domain constants are followed.
Run and interpret pnpm lint, pnpm typecheck, pnpm build, and pnpm format in a Turborepo monorepo; diagnose CI-style failures; propose minimal diffs; and deliver a fix plan, files touched, and validation commands checklist. Use when verifying CI readiness or fixing lint/typecheck/build/format failures in Turborepo repositories.
Load when reviewing code, writing code examples, or ensuring code quality. Contains coding standards for Python, JavaScript, SQL, and general best practices for publication-quality code.
Review code changes for quality, security, and maintainability. Use when user asks to 'review this code', 'check my changes', 'audit this PR', or after completing significant implementation work.
Validates code-documentation parity using tree-sitter static analysis. Use when user asks about documentation validation, doc parity, checking if docs are up to date, or whether code matches documentation.
Use when adding project dependencies. Defines dependency management rules and language-specific patterns.
Verify Java architecture rules with ArchUnit. Use when you need to define or validate package or layer boundaries, dependency constraints, slices, or cyclic dependencies in Java projects, including during refactors or when adding governance checks to CI builds.
程式碼審查最佳實踐指南。當進行 PR review、代碼審查或用戶提到「review」、「審查」時使用。
Respond to code review feedback with technical rigor. Use when (1) receiving review comments from any model (Codex, Gemini, Claude), (2) synthesizer outputs issues to fix, (3) refactor agent needs to address findings. Enforces: verify claims against codebase before implementing, push back when warranted, no performative agreement ("You're absolutely right!"), apply YAGNI ruthlessly.
