
from cc-design465
An agent skill that produces high-fidelity HTML design artifacts and interactive prototypes (slides, landing pages, UI mockups), with routing, templates, and ve
CC-Design equips an agent to act as an expert HTML designer and prototyper. The skill provides routing logic to classify design tasks, templates and references for common artefacts (slides, landing pages, interactive prototypes), and a stepwise workflow: understand, route, acquire context, define design intent, build, verify (structural/visual/design excellence) and deliver. The skill emphasizes verified outputs — no console errors, screenshot-verified rendering, and descriptive filenames — and enforces design quality checks before handoff.
Trigger this skill when a user asks to design, prototype, mock up, make it look good, or mentions Figma, slide decks, UI kits, wireframes, or presentations. Use for web landing pages, slide decks, interactive demos, and component-level HTML prototypes where visual fidelity and verified rendering are required.
Designed for agents that can run filesystem operations, spawn browser rendering (Playwright-style tools), and use Read/Write tools — suitable for Cursor, agents with Playwright integrations, or OpenClaw runner setups.
cc-design is a comprehensive HTML design and prototype skill for creating slide decks, landing pages, UI mockups, and interactive prototypes. It has 14 scripts covering video/audio processing, format conversion, PDF/PPTX export, and bundle management. The SKILL.md is exceptionally detailed with a thorough routing system, typography guardrails, anti-AI-slop rules, and verification protocols. Shell scripts properly validate inputs and show usage messages, but 10 Node.js scripts were skipped by the runner. Media scripts require ffmpeg which isn't in the audit environment.
ffmpegffproberipgrep (rg)playwright (npm)pdf-lib (npm)pptxgenjs (npm)sharp (npm)Well-crafted skill with strong design methodology. No security concerns beyond minor shell interpolation in ffmpeg filters (internal computed values, not user input in web context). Update mechanism is user-initiated, not auto-pull. Extensive reference system with clear progressive disclosure. The skill's anti-slop and verification protocols are notably thorough.