Skip to content

paolomoz/skills

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

35 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Claude Code Skills

Personal collection of 27 reusable Claude Code skills extracted from 12 production projects. Each skill encapsulates a proven pattern — from AI pipeline orchestration to brand extraction to accessibility auditing — as instructional prompts that Claude Code can follow.

Skills Catalog

# Skill Category Description
1 railway-deploy deploy Deploy monorepos to Railway with Dockerfiles, config-as-code, and GraphQL API
2 infographic-video content Transform meeting summaries into ~5 min videos with infographics and dialogue
3 infographics-generate-style content Generate infographic style definitions from reference material
4 demo-video content Build narrated demo videos from screen recordings with AI voiceover and branded slides
5 highlight-reel content Create 60-90s polished highlight reels from screen recordings using Remotion
6 screenshot-capture content Capture full-page and viewport screenshots with overlay removal
7 sse-streaming patterns SSE streaming from Cloudflare Workers with reconnection and state persistence
8 multi-model-orchestrator patterns Multi-stage AI pipelines with role-based model routing
9 cloudflare-fullstack patterns Full-stack apps on Cloudflare Workers with KV, D1, R2, Vectorize
10 session-context patterns Multi-turn session tracking with sessionStorage and KV caching
11 incremental-processor patterns Hash-based change detection for resumable batch processing
12 multi-provider-fallback patterns AI provider abstraction with automatic fallback and result synthesis
13 brand-extractor brand-design Extract brand profiles from URLs or descriptions using multi-tier AI analysis
14 brand-css-generator brand-design Generate CSS custom property overrides from brand profiles
15 figma-token-sync brand-design Bidirectional sync between CSS custom properties and Figma variables
16 design-system-extractor brand-design Extract design tokens from live websites using Puppeteer
17 generative-page-pipeline page-generation Multi-stage AI pipeline: query → intent → blocks → images → published page
18 conversational-page-builder page-generation Interactive AI agent loop for iterative page creation with Claude tool calling
19 ai-image-generator page-generation Multi-provider image generation with fallbacks, caching, and R2 storage
20 da-content-pipeline page-generation Convert HTML to DA format and upload to AEM Edge Delivery Services
21 site-auditor audit Cross-reference query indexes, sitemaps, and navigation for content gaps
22 pagespeed-audit audit Collect and analyze Core Web Vitals and PageSpeed scores
23 accessibility-auditor audit WCAG 2.1 AA compliance and content safety auditing
24 block-detector audit Detect and catalog content blocks on web pages using Puppeteer
25 content-graph data-intelligence Build semantic content graphs for navigation and recommendations
26 data-intelligence-pipeline data-intelligence End-to-end pipeline: fetch → analyze → synthesize → dashboard
27 report-hub-generator data-intelligence Generate single-page HTML report hubs with interactive navigation

Categories

deploy — Deployment & Infrastructure

Skills for deploying applications to cloud platforms.

  • railway-deploy — Railway monorepo deployment with Dockerfiles and config-as-code

content — Content Generation & Media

Skills for generating videos, images, and visual content.

  • demo-video — Screen recording + script → narrated demo with AI voiceover and branded slides (ElevenLabs + ffmpeg)
  • infographic-video — Meeting summary → infographic video pipeline (ElevenLabs + Gemini)
  • infographics-generate-style — Style definition generator for infographic production
  • highlight-reel — Screen recording → polished demo video (Remotion)
  • screenshot-capture — Automated website screenshots with overlay removal (Playwright)

patterns — Cross-Cutting Infrastructure

Foundational patterns used across multiple domain skills. These are the building blocks.

  • sse-streaming — Server-Sent Events with ReadableStream, EventSource, reconnection
  • multi-model-orchestrator — Route tasks to optimal models by role (classify → reason → generate)
  • cloudflare-fullstack — Workers + KV + D1 + R2 + Vectorize + AI bindings
  • session-context — Session tracking across multi-turn interactions
  • incremental-processor — Process large datasets incrementally with hash-based change detection
  • multi-provider-fallback — Provider abstraction with automatic fallback chains
┌─────────────────────────────────────────────────────────┐
│                    patterns (foundation)                  │
│  sse-streaming ─── multi-model-orchestrator              │
│       │                    │                             │
│  cloudflare-fullstack  session-context                   │
│       │                    │                             │
│  incremental-processor  multi-provider-fallback          │
└───────────┬────────────────┬────────────────┬────────────┘
            │                │                │
   ┌────────▼──┐    ┌───────▼───┐    ┌───────▼───┐
   │ page-gen  │    │   audit   │    │   data    │
   │ pipeline  │    │  tools    │    │  intel    │
   └───────────┘    └───────────┘    └───────────┘

brand-design — Brand & Design Systems

Extract, generate, and sync brand identity assets and design tokens.

  • brand-extractor — Multi-tier brand profile extraction (web search → AI knowledge → fallback)
  • brand-css-generator — CSS custom properties from brand profiles with WCAG AA enforcement
  • figma-token-sync — Push/pull/diff CSS tokens ↔ Figma variables
  • design-system-extractor — Puppeteer-based computed style extraction from live sites
brand-extractor ──→ brand-css-generator ──→ figma-token-sync
                          ↑                        ↑
              design-system-extractor ──────────────┘

page-generation — AI Page Creation

End-to-end pipelines for generating web pages with AI.

  • generative-page-pipeline — Query → intent → blocks → images → published page
  • conversational-page-builder — Multi-turn chat agent for iterative page editing
  • ai-image-generator — Multi-provider image generation (fal.ai, Imagen, Gemini)
  • da-content-pipeline — HTML → DA format conversion and AEM upload
generative-page-pipeline ──→ da-content-pipeline
         │                           ↑
         ├── ai-image-generator      │
         │                           │
conversational-page-builder ─────────┘

audit — Website Analysis

Audit websites for content quality, performance, accessibility, and structure.

  • site-auditor — Cross-reference indexes, sitemaps, and navigation for 11 gap categories
  • pagespeed-audit — Core Web Vitals collection and analysis
  • accessibility-auditor — WCAG 2.1 AA compliance with content safety testing
  • block-detector — Content block detection with CSS selector generation

data-intelligence — Data Pipelines & Reports

Transform raw data into structured intelligence and visual reports.

  • content-graph — Semantic content graphs for topic mapping and recommendations
  • data-intelligence-pipeline — Slack/document → AI extraction → synthesis → dashboard
  • report-hub-generator — Single-page HTML report hubs with interactive components
data-intelligence-pipeline
         │
         ├── content-graph
         │
         └── report-hub-generator

Quick Start

Using a skill

Skills are installed as a Claude Code plugin. Once installed, Claude Code matches user requests to skills via trigger keywords in the skill description.

# Install the plugin (from this repo root)
claude plugin install .

# Then just ask Claude Code naturally:
# "Set up SSE streaming for my Cloudflare Worker"
# "Extract the brand identity from https://example.com"
# "Audit this website for accessibility issues"
# "Generate a report hub from these analysis results"

Skill structure

Each skill lives in its own directory under skills/:

skills/
└── skill-name/
    ├── SKILL.md              # Instructions (required)
    ├── references/            # Supporting docs (optional)
    │   └── schema.md
    └── scripts/               # Executable scripts (optional)
        └── run.py

Adding a new skill

  1. Create a directory under skills/ with a descriptive kebab-case name
  2. Add a SKILL.md with YAML front matter (name, description with trigger keywords) and instructions
  3. Register it in .claude-plugin/marketplace.json under the appropriate plugin category
  4. Optionally add references/ for supporting documentation and scripts/ for executable tools

See skills/_template/SKILL.md for the minimal template.

Contributing

This is a personal skills collection, but the patterns are designed to be generally applicable. Each skill documents:

  • When to Use — Scenarios that trigger the skill
  • Step-by-step Instructions — What Claude Code should do
  • Code Snippets — Production-tested patterns from real projects
  • Troubleshooting — Common issues with causes and fixes
  • Cross-References — Related skills for composition

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors