skills

Assess your AI fluency using Anthropic's 4D framework (Dakan, Feller & Anthropic, 2025). Scans Claude Code sessions, Claude.ai exports, and git history, runs heuristic analysis on all messages, asks a self-assessment questionnaire, and generates a visual HTML report with scores and actionable feedback. Use when "assess fluency", "AI fluency", "fluency report", "fluency assessment", "4D framework", or "how AI fluent am I".

Use when porting a hand-coded bespoke HTML page (art-directed landing, campaign site, keynote companion, one-off microsite, designer-generated "snowflake") to Adobe Edge Delivery Services with pixel-perfect fidelity, where forcing the design into EDS's standard block library would compromise the visual intent. Also use when the source is static HTML/CSS/JS with no CMS, time-boxed to hours/days, and authoring will be done by an AI agent rather than humans in Docs/SharePoint.

Build a narrated demo video from a screen recording, a talk track script with act timelines, branded intro/outro slides, background music, and AI-generated voiceover via ElevenLabs. Use when user asks to create a "demo video", "product demo", "narrated screencast", "talk track video", or "voiceover video".

Generates a short video recap from a meeting summary. Splits content into sections, creates infographics, generates two-host dialogue audio via ElevenLabs Text-to-Dialogue API, and assembles into MP4 with background music. Use when user asks to create a "meeting video", "video recap", "video summary", or "meeting recap video".

Run automated E2E tests against a Chrome extension using Puppeteer. Use when "test extension", "e2e extension", "verify extension works", "extension testing", or "test in Chrome extension mode".

Generate artistic infographics from any topic. Runs the Sumi pipeline (analyze → structure → craft prompt → generate image) entirely within Claude Code. Use when "generate infographic", "create infographic", "sumi", "make an infographic about", or "visualize topic".

Transform incomplete or immature product input into a comprehensive AstraZeneca HCP website briefing for AEM Edge Delivery Services. Use when "build briefing", "create briefing", "write briefing", "briefing from scratch", "new site briefing", "turn this into a briefing", or "expand this into a full briefing". The output briefing is designed to be consumed directly by the eds-site-builder skill to generate the website.

Compare generated draft pages against the source briefing to catch copy errors, missing content, wrong block types, and structural mismatches. Use when "critique drafts", "check briefing compliance", "verify content fidelity", "QA against briefing", or "validate draft pages".

Generate images for web pages using multiple AI providers (fal.ai FLUX, Vertex AI Imagen, Gemini) with smart fallbacks, embedding-based caching, and R2 storage. Use when "generating AI images", "image generation", "creating hero images", or "AI-powered visuals".

Build a complete website with AEM Edge Delivery Services from scratch. Guides through project setup (from aem-boilerplate), block development, styling, content authoring, and deployment. Use this skill when the user wants to create an EDS site, set up an AEM Edge Delivery project, build EDS blocks, or start a new aem.live website. Also useful for "create a website", "new EDS project", or "AEM site from scratch".

Run comprehensive accessibility audits with test case matrices covering WCAG compliance, brand voice enforcement, adversarial testing, and content provenance tracking. Use when "accessibility testing", "WCAG audit", "a11y validation", or "content safety testing".

Detect and catalog content blocks on web pages using Puppeteer browser automation with semantic analysis, bounding box capture, and CSS selector generation. Use when "detecting page blocks", "block analysis", "content block extraction", or "page structure detection".

Generate CSS custom property overrides and complete design branches from brand profiles, ensuring WCAG AA compliance and design coherence. Use when "generating brand CSS", "creating design tokens", "brand theming", or "CSS from brand profile".

Extract comprehensive brand profiles from URLs, descriptions, or existing assets using a multi-tier AI analysis with web search fallback. Use when "extracting brand identity", "brand analysis", "brand profiling", or "creating brand profiles".

Build full-stack applications on Cloudflare Workers with KV, D1, R2, Vectorize, and Workers AI bindings. Use when building "Cloudflare Workers apps", "serverless full-stack", "Workers + D1", or "Cloudflare edge applications".

Build semantic content graphs that map relationships between pages, topics, and entities to power intelligent navigation, recommendations, and content strategy. Use when building "content graphs", "semantic networks", "content relationships", or "topic mapping".

Build interactive AI agent loops that iteratively create and refine web pages through multi-turn conversations with Claude tool calling, design selection, and real-time preview. Use when building "conversational website builder", "chat-based page editor", "AI agent builder", or "interactive page generation".

Convert generated HTML to Document Authoring (DA) format and upload with authentication, batch processing, and preview triggering for AEM Edge Delivery Services. Use when "uploading to DA", "DA content pipeline", "AEM content authoring", or "plain HTML to DA conversion".

Build end-to-end data pipelines that fetch, analyze, synthesize, and visualize intelligence from Slack channels, documents, and other sources using incremental processing and AI extraction. Use when building "data pipelines", "intelligence gathering", "Slack analysis", or "data synthesis dashboards".

Extract design tokens and component styles from live websites using Puppeteer browser automation to capture computed styles, typography, colors, and button variants. Use when "extracting design systems", "reverse-engineering CSS", "design token extraction", or "capturing website styles".

Bidirectionally sync CSS custom properties with Figma variables — push local tokens to Figma, pull Figma changes to CSS, and diff for CI drift detection. Use when "syncing Figma tokens", "Figma variables", "design token sync", or "CSS to Figma".

Build multi-stage AI pipelines that transform user queries into complete web pages through intent classification, reasoning-driven block selection, parallel content generation, and DA persistence. Use when building "page generation", "AI website builder", "content generation pipeline", or "generative web pages".

Creates a compact highlight reel video from multiple demo/screen recording videos using Remotion. Composites clips with section title cards, typewriter prompt captions, insight labels, transitions, music, and a branded poster frame. Use when user asks to create a "highlight reel", "demo video", "sizzle reel", "recap video", or "product video" from screen recordings.

Process large datasets incrementally using hash-based change detection, state persistence, and API pacing to avoid redundant work. Use when building "incremental processing", "change detection", "batch processing", or "resumable pipelines".

Orchestrate multi-stage AI pipelines that route tasks to optimal models by role — fast models for classification, reasoning models for planning, content models for generation. Use when building "multi-model pipelines", "AI orchestration", "model routing", or "LLM orchestration".

Abstract AI model providers behind a unified interface with automatic fallback, parallel execution, and result synthesis. Use when implementing "provider fallback", "multi-provider AI", "model abstraction", or "AI client switching".

Collect and analyze Core Web Vitals and PageSpeed scores across multiple URLs using the PageSpeed Insights API and RUM data aggregation. Use when "running PageSpeed tests", "Core Web Vitals audit", "performance benchmarking", or "CWV analysis".

Generate single-page HTML report hubs that aggregate multiple analysis reports with interactive navigation, view toggles, and consistent theming. Use when building "report dashboards", "analysis hubs", "HTML reports", or "data visualization pages".

Capture full-page and viewport screenshots of websites using Playwright with overlay removal, cookie consent handling, and comparison modes. Use when "capturing screenshots", "website screenshots", "visual testing", or "page capture".

Track user session context across multi-turn interactions using browser sessionStorage and server-side KV caching with TTL. Use when implementing "session tracking", "conversation context", "multi-turn sessions", or "user journey tracking".

Audit websites by cross-referencing query indexes, sitemaps, and navigation to identify content gaps, stale pages, missing metadata, and quality issues. Use when "auditing a website", "finding content gaps", "site quality audit", or "content inventory analysis".

Implement Server-Sent Events streaming from Cloudflare Workers to browser clients with reconnection, state persistence, and progress tracking. Use when building "SSE streaming", "real-time updates", "server push", or "event streaming".

Generates a new infographic style definition from reference material (images, PPTX files, URLs, or verbal descriptions). Analyzes the source, extracts design DNA, and writes a style markdown file. Use when user asks to "create a style", "new infographic style", "generate style", "add a style from", or "extract style from".

Deploy a monorepo to Railway with multiple services using Dockerfiles, config-as-code, and the Railway CLI/GraphQL API. Use when deploying apps to Railway, setting up Railway projects, or fixing Railway build failures.