🧩 Skills Hub

Generates an interactive, self-contained HTML travel itinerary — light/dark theme, animated flight routes, day-by-day schedule, hotel cards, SVG maps, budget, and persistent checklists. The output looks like a premium app, not a document. It starts with a conversational discovery flow to gather real details about your trip.

Complete folder: skills/travel-itinerary/ (SKILL.md + README.md + DESIGN-REFERENCE.md + example-tokyo-trip.html).

Fixes layout, CSS, and DOM problems on the frontend by prioritizing live visual feedback in the browser before writing any changes to the source code. Use it when you want to test CSS/HTML changes, fix clipped elements, or adjust layouts visually. It works in a sandbox branch for safety and only persists changes after your approval.

Reviews n8n workflows for errors, inefficiencies, and missing best practices—as a senior automation architect doing a clear, actionable code review. Accepts complete JSON, partial config, a text description, or a screenshot, and runs a five-category review framework.

Designs the ideal RAG system for your data: analyzes structure, format, relationships, and scale to recommend the right architecture—and then delivers an implementation plan with starter code. It triggers on any request about RAG, vector/semantic search, chunking, or “chat with your documents,” even without the word “RAG.”

Specialist in traditional SEO and AI search optimization (AEO). Conducts site audits, schema markup, Core Web Vitals, llms.txt, and citation strategy — using bash, file creation, and web fetch to deliver real results, not just advice. Covers visibility in ChatGPT, Perplexity, Gemini, and AI Overviews.

Full-stack client acquisition and delivery engine: combines competitive research, brand strategy, a customized sales pitch, operational agent architecture, and self-evolving workflow capture in a single system. Takes a business from cold prospect to fully operationalized client through the PROSPECT → PITCH → BRAND → OPERATE → CAPTURE phases.

Gives your AI agents a shared brain that works across machines, tools, and frameworks: save a fact in Claude Code on your laptop and retrieve it from another agent on the server. Includes the ritual /sessionend, which reflects on the session, saves a structured summary in the Shared Brain, and updates local memory.

Qualifies, scores, and filters leads using Apify Google Search. Classifies businesses that sell their own products as brand and scores each one across 3 dimensions—product development capability, data maturity, and brand maturity—from a leads CSV. Includes the Python script that performs the enrichment.