Eldergenix

Use for product design, UI/UX, branding, high-fidelity web/mobile/desktop interfaces, HTML prototypes, design systems, frontend implementation, and design quality work.

Use to create, repair, or enforce design tokens, semantic surfaces, light/dark themes, spacing, radius, typography, component variants, and reusable UI primitives.

Use for empty states, no-data UI, AI output streaming, thinking steps, loading states, skeletons, progressive disclosure, and conversational product surfaces.

Use after design or implementation to package handoff notes, refactor UI code, split god components, document tokens, and prepare maintainable frontend deliverables.

Use for React Native, Expo SDK 56, NativeWind, iOS/Android UI, mobile UX flows, touch targets, safe areas, and mobile design-system implementation.

Use for interaction design, animations, progressive blur, animated borders, liquid metal accents, transitions, gesture feedback, and motion quality review.

Use before major product, UI/UX, branding, or prototype work to clarify goals, audience, constraints, design system, variations, and deliverables.

Use to review UI/UX quality, identify design inconsistencies, accessibility issues, visual polish defects, implementation debt, and production-readiness gaps.

SoundService, SoundGroup mixer, spatial audio, 3D positioned sound, music systems, SFX patterns.

3D builds via MCP, CSG operations, spatial coordination, Part properties, map design.

Lighting service, Atmosphere, post-processing (Bloom, ColorCorrection), day/night cycle.

Luau type system: annotations, generics, narrowing, inference philosophy, sealed/unsealed tables, exports, and Roblox-aware typing.

Constraints, VehicleSeat, ragdoll, projectiles, Attachment patterns, network ownership.

Studio MCP server tools, execute_luau, script_read/multi_edit, reliability patterns, workflows.

Route approved URLs, HTML, public Figma JSON, images, PDFs, PSD, AI, EPS, or PS through a configured any-to-design MCP; extract CSS-backed design systems, generate agent-ready artifacts, and bridge editable scenes to or from Figma. Use when users ask to convert a website or design source, reconstruct a design system, create editable Figma layers, compile design tokens or components for web, mobile, dashboards, presentations, PDFs, or documents, or explicitly use hosted code.to.design/anything.to.design APIs.

Single-vendor multi-agent coding pipeline with tiered model routing: your main session orchestrates and plans, a cheap Claude coder (e.g. Sonnet) implements standard items and a strong one (e.g. Opus) takes the hard ones, a strong-model subagent reviews adversarially, every finding passes an evidence-based TRIAGE before it can loop, and an orchestrator acceptance gate audits criteria-vs-evidence before merge. Use when the user wants an orchestrated implement→review→merge lifecycle WITHOUT a second vendor ("all-Claude", "anthropic-only", "no codex", or no Codex CLI installed), wants cost-tiered delegation (cheap coder on crisp briefs, strong coder on hard parts), or says "orchestrate this" in a Claude-only environment. Requires: git repo + a subagent-capable harness. If the OpenAI Codex CLI IS available and the work is high-stakes, prefer the sibling cross-vendor-orchestrator skill (uncorrelated review beats same-family review).

Maintain a personal knowledge base as an LLM Wiki (the Karpathy pattern): the human curates immutable raw sources and asks questions; the agent owns a wiki layer of summary/entity/concept pages, a catalog index, an append-only operation log, and a deterministic integrity gate that blocks bad commits. Use when the user wants a "second brain", personal wiki, or compounding notes vault; says "ingest this article/paper/link into my notes", "what does my knowledge base say about X", or "lint/health-check my vault"; or wants agent memory that a human can read and browse (Obsidian-compatible markdown + wikilinks). Includes the three operations (ingest / query / lint), page and provenance conventions, a runnable pre-commit check script, and scheduled-automation patterns.

Multi-agent coding pipeline with cross-vendor adversarial review: your main session orchestrates and plans, a Claude subagent (strongest coder, e.g. Opus) implements in an isolated tree, OpenAI Codex (GPT-5.6 Sol) adversarially reviews AND applies fix rounds, and nothing merges until an independent APPROVE plus an orchestrator merge gate that re-runs the artifact end to end. Use when the user wants orchestrated/delegated implementation with independent review ("orchestrate this", "cross-vendor review", "have GPT review Claude's code", "adversarial pipeline", "implement with review loop"), or hands over multi-file or risky work that deserves more than a single-model pass. Requires: git repo, a subagent-capable harness, and the OpenAI Codex CLI (`codex login` done). Not for trivial single-file edits.

Use when the user asks you to build, style, theme, refine, audit, or recreate UI — or uploads a screenshot to rebuild. Reads DESIGN.md as a closed token layer, detects the project's framework (Tailwind v4/v3, ShadCN, MUI, Radix, Geist), and enforces production design rules (tokens only, full state matrix, WCAG 2.2 AA, responsive across 320→1920px, reusability, 300 LOC hard limit, 44×44 touch targets, continuous UI memory) via PostToolUse hooks.