Skill UI

This skill reviews generated or changed documentation before release, verifying claims against source code to prevent hallucinations. It checks API references, docstrings, and code samples for accuracy. Use it reactively after documentation updates or proactively before writing to ensure technical correctness and version alignment.

This skill enables reviewing defined code diffs from correctness, authorization, and performance perspectives. It supports bounded multi-agent review when authorized, ensuring findings are reproduced centrally. Users must pin repository paths and intended behavior. It does not install orchestration engines but provides a structured procedure for code quality assurance and security validation without overstepping user authority.

Invariant-Guard enforces a correctness-first approach to algorithm design. Instead of writing code and hoping it works, this skill mandates defining the function contract, loop invariants, termination arguments, and all potential edge cases *before* implementation. It is essential for complex algorithms where subtle bugs (like off-by-one errors or missed post-conditions) are common, ensuring mathematically rigorous and verifiable code.

This skill identifies logic bugs in code files or functions using semi-formal execution tracing. It follows a structured process: Premises, Trace, Divergence, Trigger, and Remedy. Ideal for reviewing code when specific failures aren't named but correctness is suspected.

Luna is an automated code reviewer focused on correctness, security, and reliability. She evaluates code against blueprints and checklists, identifying critical vulnerabilities like injection flaws, authentication bypasses, and race conditions. Luna acts as a quality gate, ensuring no unresolved high-risk issues proceed to QA or deployment. She excludes style preferences, focusing only on measurable risks to maintainability and safety.

A comprehensive engineering stack for designing and implementing robust, production-grade payment gateways for Software-as-a-Service (SaaS) platforms targeting the Pakistani market. It covers integrating major local providers like JazzCash and Easypaisa, handling complex payment flows (PKR billing, webhooks, idempotency), and performing essential financial reconciliation and auditing processes. This guide ensures high correctness and reliability in B2B/SaaS billing operations in Pakistan.

Coordinates a bounded code review of a defined diff from correctness, authorization and performance perspectives, delegating to independent agents only when authorized and reproducing important findings centrally.

This specialized skill reviews user interface animations, motion patterns, and interactive code (CSS, Framer Motion, etc.) against a strict, high-craft standard. It evaluates not only functionality but also performance, accessibility, physical correctness, and adherence to motion design principles. Use it when reviewing diffs related to transitions, keyframes, or gestures to ensure the motion feels right, not just that it runs.

This skill enforces a high standard for all generated code, prioritizing correctness, clarity, and idiomatic patterns across all languages. It helps minimize code bloat by removing unnecessary abstractions and verbose patterns, while also optimizing the agent's output by using targeted diffs and eliminating excessive conversational overhead. Use this for all code generation, refactoring, or review tasks.

Fine-tune LLMs on a single consumer GPU with Unsloth Core. Covers VRAM sizing, LoRA/QLoRA, GRPO/DPO post-training, chat-template correctness, and GGUF export.

A specialized tool for auditing WordPress plugins, themes, and blocks before deployment. It checks for security vulnerabilities like SQL injection, XSS, and improper escaping, ensures API correctness, and enforces internationalization standards. Use it to review generated code or perform security assessments on WordPress environments.

A five-axis code review checklist covering correctness, security, readability, performance, and test coverage to produce actionable comments instead of style nitpicks. Use for reviewing code, diffs, or pull requests.