A collection of specialized GitHub Copilot CLI skills that extend Copilot with structured, domain-specific workflows.
Generates high-impact, ATS-optimized resumes in reverse chronological format from raw, messy, or unstructured user input. Takes anything — LinkedIn dumps, project notes, loose bullet points, or a wall of text — and produces a polished, recruiter-ready, compilable LaTeX document tailored to a specific role and company. Includes keyword alignment from job descriptions, role-level calibration, and an iterative review loop until the user is satisfied.
A structured three-stage workflow for co-authoring documentation. Guides users through context gathering, iterative refinement, and reader testing to produce polished technical docs, proposals, specs, and decision records.
Creates distinctive, production-grade frontend interfaces with high design quality. Applies bold aesthetic directions — minimalism, brutalism, retro-futuristic, and more — to build web components, landing pages, dashboards, and React apps that avoid generic "AI slop" patterns.
A five-stage structured workflow for debugging Linux kernel issues. Handles panics, oops, lockups, and memory corruption by classifying the failure, collecting artifacts, identifying root causes, and recommending the right debugging tools (ftrace, kprobe, kgdb, dynamic debug).
An eight-stage workflow for writing Linux kernel device drivers from scratch. Covers device type identification, subsystem architecture review, minimal skeleton implementation, device tree/ACPI bindings, probe/remove lifecycle, power management, and idiomatic error handling — always producing compilable code.
Reviews Linux kernel patches with the rigour of an experienced maintainer. An eight-stage process covering commit message quality, coding style compliance, API correctness, locking and memory safety, performance, and maintainability — with severity-rated findings (BLOCKER → NITPICK).
Guides researchers from problem definition through publication-ready output. A ten-stage workflow covering literature review, hypothesis, methodology, results, paper structure, citation management (BibTeX), and output formats (LaTeX, PDF-ready, Word).
A meta-skill for creating and improving other skills. Supports iterative development with draft writing, eval test prompts, quantitative benchmarking, and skill description optimization for accurate triggering.
Designs production-grade software systems by emulating an experienced systems architect. An eight-stage workflow covering requirement clarification, domain modeling, architecture strategy, component decomposition, technology selection, failure analysis, and full output artifacts including Mermaid diagrams.
A styling toolkit with ten pre-built professional themes (Ocean Depths, Sunset Boulevard, Modern Minimalist, Midnight Galaxy, and more) plus on-the-fly custom theme generation. Applies cohesive color palettes and font pairings to slides, docs, reports, and web artifacts.
A nine-stage workflow for deeply analyzing any Linux kernel subsystem at engineer level. Walks through scope, architecture, core data structures, locking rules, key execution paths, real debugging scenarios, and how to submit patches — leaving developers ready to read source, write patches, or debug real bugs.
Builds elaborate, multi-component Claude.ai HTML artifacts using React 18, TypeScript, Tailwind CSS, and shadcn/ui (40+ components). A structured workflow — initialize, develop, bundle to a single self-contained HTML file — for complex artifacts requiring state management or routing.