Skill UI

This powerful skill provides an interactive debugger interface, allowing developers to set breakpoints, step through source code line by line, inspect live variable states, and trace the entire call stack. It is essential for diagnosing complex bugs, understanding unexpected program flow, or determining the root cause when standard logging fails. It supports various languages via the DAP protocol.

This skill guides users through an in-depth interview process to achieve maximum project context completeness (100% "Gold Code") for their .faf file. By systematically asking questions about project goals, architecture, and technology stack, it ensures all critical context slots are filled, making the project fully structured and highly readable by AI systems.

This skill defines a framework-agnostic, scalable architecture pattern for React and React Native applications. It mandates feature modules, strict state separation (server vs. UI), and disciplined import rules to create predictable, maintainable, and highly scalable codebases, regardless of the chosen framework (Next.js, Remix, etc.).

This discipline establishes a portable, framework-agnostic layer for ensuring type safety at the network edge of any React or React Native application. It mandates a single, typed API client to wrap all network fetching, guaranteeing that raw wire JSON is immediately parsed into trusted, domain-specific types or standardized error objects before entering the application state. This pattern drastically reduces runtime errors and enhances code robustness.

This skill is designed to accurately convert complex, Hugo-managed documentation sites and content into standard Markdown format. It goes beyond generic knowledge by inspecting the entire local Hugo repository structure, including `hugo.toml`, custom shortcodes, layouts, and content files. It preserves the original content's semantic meaning while materializing Hugo-specific constructs into clean Markdown.

This skill generates a detailed, step-by-step, interprocedural execution trace for a given piece of code and specific input scenario. It resolves name ambiguities, tracks type transitions, and highlights non-obvious behaviors (like implicit coercions or side effects). Use it when a user is confused about how code behaves, allowing for a deep dive into the actual execution path to clarify assumptions versus reality.

This skill provides an autonomous, repository-wide pipeline for auditing and fixing complex logical inconsistencies. It systematically executes stages including health checks, deep review, locating divergences, generating fixes, and performing diff-verification until the codebase is deemed clean. Ideal for maintaining high code quality and resolving deep, systemic bugs across entire projects.

This skill is designed to identify subtle and complex logic bugs within a single code file or function. It employs a semi-formal execution tracing methodology, systematically breaking down the suspected bug into five stages: Premises (expected state), Trace (execution path), Divergence (actual deviation), Trigger (reproduction case), and Remedy (fix). It is ideal for scenarios where users suspect code is wrong but cannot specify a concrete failure point.

StyleSeed Design Review analyzes UI/frontend code (React/Tailwind/HTML) against a detailed design rubric. It scores the interface (0-100) across coherence, color discipline, hierarchy, and polish. The tool pinpoints specific violations—such as mixed corner radii or non-semantic colors—and generates a prioritized, actionable list of fixes, ensuring the final product looks professional and polished, not generic or 'AI-generated.'

This skill analyzes your project's current StyleSeed setup, detects outdated engine files, and safely updates the core components, skills, and rules. It follows a structured, non-destructive process, ensuring that user-defined code, such as theme.css and custom components, remains untouched while applying necessary upgrades.

Specialized skill to rewrite and humanize AI-generated natural language files (like notes or documentation). It removes common 'AI-isms' such as sycophantic openings, stock vocabulary, over-explaining, and excessive signposting. Users can choose between a fast, deterministic regex cleanup or a deep LLM rewrite (default mode) using Claude to enhance burstiness and match a natural human voice, all while strictly preserving code blocks, URLs, and headings.

Rewrites overly formal or vague code review comments generated by AI or corporate tools. It cuts out filler phrases like "I noticed..." or "It might be worth considering...", transforming them into direct, actionable feedback. The output uses a clear format (Location: Severity: Observation. Fix.) that sounds like it came from a helpful, experienced teammate, significantly improving the quality and reception of PR feedback.