Claude Skills by SkyyRoseLLC

Plans YouTube channel strategy with content pillars, upload schedule, thumbnail guidelines, and growth milestones. Use when launching or restructuring a YouTube channel for business growth.

Have two genuinely different models (a Claude model, e.g. Fable, and Codex/OpenAI) debate a plan before anything gets built — one drafts, the other challenges, capped at 3 rounds, then the strongest coding model actually executes the converged plan and the planner reviews the real result. Use for any task where a wrong plan is expensive to discover late (rigging/animation work, architecture decisions, migrations) — not for small, obviously-correct changes.

Node.js / TypeScript backend patterns: REST API design, repository pattern, middleware, caching, and error handling for Express and Next.js API routes. (Python/FastAPI service layers live in the fastapi-patterns skill.)

Use to work the way the Claude Code team works — parallelize aggressively across worktrees and subagents, plan before non-trivial work, verify everything, and reach for /loop, /batch, /simplify, and effort:max on the right tasks. The power-user operating posture, distilled.

Database migration best practices for schema changes, data migrations, rollbacks, and zero-downtime deployments across PostgreSQL, MySQL, and common ORMs (Prisma, Drizzle, Kysely, Django, TypeORM, golang-migrate).

Use when adding, upgrading, or resolving Python/Node dependencies — declare every import, cap majors to stop silent drift, and keep the resolver consistent. Prevents undeclared-import and silent-major-bump defects.

Always-on tool discipline and production output enforcement. Fires every session. Enforces: no file re-reads, batched parallel tool calls, anatomy.md check first, one targeted search, zero TODOs/mocks in delivered code, verify-before-claim anti-hallucination, HTML for structured agent docs. Complements token-aware-behavior (context window) and caveman (tone) — this one governs tool efficiency and output quality.

Use when you have a written implementation plan to execute in a separate session with review checkpoints

Use when FastAPI async patterns for building high-performance APIs. Use when handling concurrent requests and async operations.

Use for symbol-precise code navigation on the monorepo — go-to-definition, find-references, rename, hover-types via the installed LSP servers instead of grep guessing. Answers from the real symbol graph, not text matching.

Use at the start and end of every task — read the canonical memory before acting and write the outcome after. Unifies OpenWolf files and claude-mem so knowledge persists across sessions instead of being re-derived.

Use when the user asks to prototype a feature, mock something up, explore options, see variations or alternatives, or says "what could this look like" — any request to trial a feature before committing to one approach.

Use when structuring how you approach any task — select the reasoning technique that fits the task type (chain-of-thought, ReAct, few-shot, tree-of-thoughts, self-consistency, structured output) to produce production-quality work, always subordinate to verification.

Python best practices, security patterns, type hints, async patterns, and Pythonic idioms for FastAPI/Django projects.

Redis data structure patterns, caching strategies, distributed locks, rate limiting, pub/sub, and connection management for production applications.

Use after any correction, mistake, discovery, or fix — capture the lesson durably into cerebrum + buglog so the same discovery is never repeated. Closes the improvement loop that makes the team better across sessions.

Handoff router + memory/self-heal wiring for skyyrose-core. Use when an engineering task needs verification or design work, and to read/write the suite's shared memory substrates.

Use before caching, quoting, or acting on ANY product, imagery, or brand fact — resolve it from the canonical source registered in SOT.md, never from memory or a filename. Prevents second-copy drift and wrong-fact defects.

Recover gracefully when an Agent tool dispatch stalls, times out, or returns mid-edit. Detect via duration > 300s + partial-edit signature, locate agent's last write target, complete inline. Use when an Agent invocation returns a non-completion message, when the agent's last sentence is mid-thought, or when you suspect token budget hit.

Always-on context window management for Claude chat sessions. Activates automatically on every conversation — no trigger phrase needed. Instructs Claude to proactively monitor conversation length, compress history at 75% depth, emit structured handoff documents at 88%, and hard-stop at 95%. Use this skill whenever a conversation is running long, involves multi-step engineering tasks, contains large code blocks or tool results, or the user mentions "context", "losing track", "starting fresh", ...

Offers the user an informed choice about how much response depth to consume before answering. Use this skill when the user explicitly wants to control response length, depth, or token budget. TRIGGER when: "token budget", "token count", "token usage", "token limit", "response length", "answer depth", "short version", "brief answer", "detailed answer", "exhaustive answer", "respuesta corta vs larga", "cuántos tokens", "ahorrar tokens", "responde al 50%", "dame la versión corta", "quiero contro...

Use before any git commit/stash/history operation in a repo that multiple Claude sessions may share — never rewrite another session's HEAD, never pop a shared stash. Prevents cross-session commit corruption.

Design, implement, and audit inclusive digital products using WCAG 2.2 Level AA

Generate AI images with GPT-Image-2, FLUX, Gemini, Grok, Seedream, Reve and 50+ models via inference.sh CLI. Models: GPT-Image-2, FLUX Dev LoRA, FLUX.2 Klein LoRA, Gemini 3 Pro Image, Grok Imagine, Seedream 4.5, Reve, ImagineArt. Capabilities: text-to-image, image-to-image, inpainting, LoRA, image editing, upscaling, text rendering. Use for: AI art, product mockups, concept art, social media graphics, marketing visuals, illustrations. Triggers: flux, image generation, ai image, text to image,...

Diagnose and fix a WordPress/theme issue that resists a single-shot patch, by running parallel independent fix hypotheses through adversarial verification instead of trusting one agent's self-graded "fixed it." Use for regressions escalated by self-healing-theme-loop (recurrences >= 2, or root cause genuinely uncertain), or any skyyrose.co bug where a first fix attempt didn't hold up under independent re-check. Not for routine, known-signature regressions -- theme-heal-doctor's single-shot pa...

SkyyRose gpt-image-2 prompt composer. Given a SKU (or freeform request), collection, presentation mode, and view, assembles a production-ready gpt-image-2 edit-mode prompt in the exact grammar of scripts/oai_render/prompt.py and emits the gpt-image-2 output contract JSON. Never fires a paid render without the STOP-AND-SHOW manifest and explicit "y" from the user.

Video and audio generation via fal.ai MCP. Covers text/image-to-video (Seedance, Kling, Veo 3), text-to-speech (CSM-1B), and video-to-audio (ThinkSound). For product or still images use the ai-image-generation skill (OpenAI Image 2 is the canonical SkyyRose engine) — fal.ai image models are not used for product imagery. Use when the user wants to generate videos or audio with AI.

Set an ECC-specific frontend design direction for production UI work. Use when building or improving websites, dashboards, applications, components, landing pages, visual tools, or any web UI that needs stronger product-specific design judgment.

Build production-grade frontends grounded in the user's existing codebase, framework, and design system. Read first, design second. Defaults: accessible (WCAG 2.2 AA), performant (Core Web Vitals budgeted), responsive (mobile-first). Picks aesthetic register from product context (calm-editorial / dense-utility / expressive-brand) rather than impulse. Includes the Claude calm-editorial design vocabulary as one register among many, not as the default. Replaces the upstream frontend-design skill...

Specialized architect for building augmented reality, 3D, and immersive shopping experiences — across ANY output format (React JSX, standalone HTML/CSS/JS, WordPress/WooCommerce PHP, Python backends, or any other language). Use this skill whenever the user mentions: 3D product viewers, virtual try-on, AR fitting rooms, immersive brand showrooms, 3D collection environments, interactive drop pages, Three.js, React Three Fiber, A-Frame, model-viewer, WebXR, GSAP cinematic scroll, Lottie animatio...

Expert knowledge in building interactive, dynamic web experiences with React, Vue, Svelte, animations, and user interactions. Triggers on keywords like "interactive", "animation", "dynamic", "react", "vue", "svelte", "gsap", "framer motion", "scroll animations", "parallax", "hover effects", "transitions".

Five-agent design-elevation pipeline for luxury HTML mockups. Audits a design-reference HTML file against brand canon, identifies polish gaps, generates a fix manifest, executes infrastructure changes via Senior Developer, and verifies the result. Use when refining a design mockup from "decent" to "production-grade luxury". Trigger phrases: "elevate this mockup", "luxury polish pass", "5-agent design pipeline", "/luxury-mockup-pipeline".

MANDATORY before rendering, creating content, or editing skyyrose.co — eyes-on verify every product image is the CORRECT garment for its SKU by reading the actual pixels (vision), never the filename or manifest. Prevents wrong-garment imagery, the #1 recurring defect.

Build a throwaway prototype to answer one question before committing — in ANY domain. Routes between three branches — a runnable terminal app for state/business-logic questions, several radically different UI variations toggleable from one route, or several variant generated artifacts (image, video, 3D, audio, copy) compared side-by-side. Use when the user wants to prototype, sanity-check a data model or state machine, mock up a UI, explore image/render/lookbook directions, compare engines, e...

Handoff router for skyyrose-design. Use when a design/build task needs copy, backend, or verification owned by another SkyyRose plugin — tells you when and where to hand off.

Use after editing any SkyyRose theme CSS/JS — production serves the .min build, so every source edit must rebuild .min and re-verify BOTH source and .min. Prevents shipping stale minified assets.

Extracts reusable prompt elements from any input source — SkyyRose dossiers, gpt-image-2 prompts, brand docs, and design templates — into a structured element library aligned with the SkyyRose brand canon and gpt-image-2 output grammar. Supports SKU-aware dossier ingestion, brand-canon validation, human-review diff gating, and manifest export for design-master consumption.

Review WooCommerce code changes for coding standards compliance. Use when reviewing code locally, performing automated PR reviews, or checking code quality.

Receive and verify WooCommerce webhooks. Use when setting up WooCommerce webhook handlers, debugging signature verification, or handling e-commerce events like order.created, order.updated, product.created, or customer.created.

WordPress and WooCommerce development guidelines with PHP best practices, security standards, and extensibility patterns

Automate WordPress and WooCommerce operations via REST API. Use when managing products, orders, content, or site configuration programmatically. Triggers on requests to create/update products, manage inventory, generate Elementor layouts, handle WooCommerce orders, or automate any WordPress operations. Includes Elementor JSON generation patterns for skyyrose.co.

Multi-platform content distribution across X, LinkedIn, Threads, and Bluesky. Adapts content per platform using content-engine patterns. Never posts identical content cross-platform. Use when the user wants to distribute content across social platforms.

Brand identity guide for SkyyRose luxury streetwear. Use when creating content, designs, marketing materials, product descriptions, or any creative work for SkyyRose. Triggers on references to SkyyRose, the collections (Black Rose, Love Hurts, Signature, Kids Capsule), Oakland streetwear, luxury fashion content, or requests for brand-aligned outputs. Essential for maintaining brand consistency across all touchpoints.

Index hub for the SkyyRose social media branch — routes to 27 specialized social skills (content formats, strategy, community, influencer/PR, visual briefs) and the Elite Studio social venture. Use to find the right SkyyRose social skill, or for the at-a-glance brand-voice quick reference.

Complete Klaviyo email marketing system for SkyyRose luxury streetwear. Welcome sequences, pre-order drop launches, abandoned cart recovery, post-purchase flows, and seasonal campaigns. Use when creating any email content, flows, or sequences for skyyrose.co.

Full creator-program lifecycle for SkyyRose — from discovery profiling and personalized outreach (DM <150 words + email with subject) through a 3-touch follow-up sequence (day 0 / day 5 / day 15, hard stop), compensation tiers (nano gifting+affiliate through macro flat+usage), collab agreement outline, campaign brief, content approval workflow (max 2 rounds / 48 h turnaround), performance tracking (unique code + UTM), post-campaign analytics (day-7 screenshot), and anti-pattern recovery. Trig...

End-to-end product drop orchestration for SkyyRose. Coordinates the full launch timeline from T-30 to T+7 across product setup, email sequences, ad campaigns, social content, and influencer seeding — with mandatory STOP-AND-SHOW gates before any paid spend, Klaviyo sends, WooCommerce writes, or media uploads. Use when planning any product drop, collection launch, restock, or seasonal event.

Handoff router for skyyrose-market. Use when a marketing task needs work owned by another SkyyRose plugin (imagery, pages, backend, or verification) — tells you when and where to hand off.

Paid media playbook for SkyyRose luxury streetwear — Meta, Google, TikTok campaigns. Covers campaign architecture (verified API field names), pixel + CAPI server-side events, Merchant Center feed with preorder fields, UTM attribution, audience strategy, creative copy, and brand-canon enforcement. Use when building, auditing, or optimizing paid campaigns.

Produces a photographer-ready brief for any SkyyRose photography session — product/e-commerce shots (PDP catalog: hero front, angle, back, detail, flat-lay, on-body, lifestyle) AND brand/lifestyle/founder/campaign/BTS sessions. Covers per-collection visual direction, lighting, shot-list tables, platform crop specs (WooCommerce PDP, IG, Stories/Reels, Pinterest, web hero, email), file naming, styling, AVOID list, and Elite Studio handoff. Merges both product-photography and brand-photography b...