Skill UI

This framework enforces high output quality, intellectual honesty, and continuous improvement across all AppSec agents. It mandates structured processes, including self-critique loops, anti-rationalization guards, and self-reflection quality gates (≥8 scoring). Applicable during code review, SAST, SCA, and threat modeling, it ensures rigorous and verifiable security analysis.

This comprehensive guide serves as a decision flowchart for developers using Claude Code. It details when to use specific slash commands (/plan, /compact, /loop), select specialized agents (planner, code-reviewer, build-error-resolver), or invoke workflow skills (e2e-testing, tdd-workflow). It helps users select the optimal tool combination for complex tasks across the entire software development lifecycle, including feature planning, bug fixing, testing, and architectural design.

This tool analyzes and refactors code to detect and correct anemic domain models. It enforces rich domain modeling patterns (Entities, Value Objects, Aggregates, Domain Events), improving encapsulation and business logic ownership. Use it when validating, reviewing, or refactoring core domain logic to ensure adherence to Domain-Driven Design (DDD) principles.

This skill performs comprehensive, feature-level integration validation after all associated development tasks are implemented. It goes beyond checking individual tasks to ensure cross-task consistency, full test suite passage, and complete specification coverage. It validates data flow, architectural adherence, and performs critical security scans for residual issues (e.g., TODOs, hardcoded secrets), providing a holistic GO/NO-GO assessment for deployment. Use this skill as a final quality gate to verify that the completed feature set respects the designed boundaries and dependencies end-...

This tool assists engineers in transforming vague work logs into structured, evidence-backed impact statements for professional documentation. It uniquely mines Copilot CLI session logs, git commits, and GitHub PRs to reconstruct forgotten work. It enforces a mandatory 'Action → Result → Evidence' impact contract, making it indispensable for preparing performance reviews, promotion packets, and self-assessments.

This skill analyzes your entire codebase to ensure it is optimized for AI contributions and developer tooling. It automatically generates essential configuration files such as `AGENTS.md`, Copilot instructions, CI workflows, and customizable issue templates. It learns from your existing PR review patterns to tailor the setup precisely to your project's stack. Use it when you need to prepare a repository for advanced AI-assisted development.

This advanced code reviewer analyzes your codebase through the principles of 12 classic software engineering books (e.g., Clean Code, DDIA, Mythical Man-Month). It detects conceptual flaws, design smells, tight coupling, missing abstractions, and architectural risks—issues that standard linters and basic AI tools often miss. Essential for major refactoring and pre-design validation.

Logic Lens is an advanced AI skill that performs deep, formal logic-driven code review. Unlike standard linters, it analyzes code for complex issues like race conditions, security vulnerabilities (SQL injection, XSS), type safety flaws, and algorithmic errors across 9 critical risk categories. It is ideal for pre-merge PR reviews, auditing sensitive code paths, and refactoring complex business logic.

Cavecrew provides a set of specialized subagents (investigator, builder, reviewer) designed for efficient code development delegation. Unlike vanilla agents, Cavecrew compresses output results, drastically reducing the token usage injected back into the main context. This allows users to maintain context over much longer, complex coding sessions, effectively preventing context exhaustion during large-scale refactoring, debugging, or multi-step feature development.

Performs a comprehensive external audit of a deployed repository's state, going beyond simple code review. It checks for critical production gaps, including RLS misconfigurations, webhook idempotency issues, secret exposure, deployment health, and prompt injection vulnerabilities. Ideal for pre-launch gates and ensuring code stability before going live.

This skill performs a comprehensive post-generation visual polish on existing slide decks (PPTX/Beamer). It utilizes per-page Codex review against a specified reference PDF to ensure consistent visual weight, typography, and layout across all slides. It fixes font scaling, text overflow, and structural drift without altering core content, ensuring a professional and cohesive presentation.

MockHunter is an advanced skill designed to audit live web pages, providing a comprehensive reality check on visible data. It analyzes every element to determine its true source: whether it is REAL (from the database), MOCK (simulated data), LLM-generated, HARDCODED, BROKEN (broken endpoints), or UNKNOWN. Ideal for auditing AI-generated UIs or reviewing proof-of-concept MVPs before customer sign-off.