Use when user seeks domain-specific expertise shaped by a practitioner's methodology, not just information. The user wants their work approached from a recognized expert's perspective - how would [specific expert] solve this problem, write this code, design this system.
Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, artifacts, posters, or applications (examples include websites, landing pages, dashboards, React components, HTML/CSS layouts, or when styling/beautifying any web UI). Produces one committed, polished design direction; for exploring multiple contrasting design directions, use vs-design-diverge instead.
Automated PR review-fix loop: request bot reviews, poll for responses, analyze comments, fix code, internal review, push, repeat until no critical issues remain. Use when user says 'autofix PR', 'fix PR reviews', 'PR review loop', or '/git-pr-autofix'.
Use when user wants coaching-style resume improvement rather than simple AI rewriting. Guides users through a structured discovery process to uncover hidden strengths and create genuine differentiation for a target job posting.
Create non-generic, high-entropy frontend interfaces using Verbalized Sampling (VS) to break mode collapse and maximize design creativity. Use when the user wants to explore multiple distinct design directions, compare aesthetic options before committing, or explicitly asks to avoid generic AI-looking design (e.g. "make it less generic", "show me bolder options", "explore different design directions"). For a single polished implementation without direction exploration, use frontend-for-opus-4-5 instead.