Noureddine RAMDI / agent-skills by sleekdotdesign: modular AI agent skills for programmatic mobile app design

Created Mon, 04 May 2026 10:23:02 +0000 Modified Mon, 06 Jul 2026 15:16:10 +0000
853 Words
github-stars ai agent-skills mobile-app-design cli api nodejs

sleekdotdesign/agent-skills

The landscape of AI coding assistants is evolving beyond static capabilities. Instead of building everything into a monolithic model, AI agents are gaining installable, domain-specific skills that let them extend functionality on demand. sleekdotdesign’s agent-skills repository embodies this shift by providing a modular skill that lets AI agents design mobile apps using Sleek’s API — programmatically creating screens and managing projects.

What agent-skills provides and how it works

At its core, this repository offers a single, focused “design-mobile-apps” skill for Sleek, an AI-powered mobile app design tool. The skill enables AI agents to interact with Sleek programmatically, automating parts of the design workflow like creating screens and managing projects through API calls.

The architecture follows the emerging agent skills pattern: capabilities are packaged as discrete, installable modules that AI coding assistants can discover and add dynamically. This skill is distributed via the npx skills CLI tool, which handles installation and integration into the agent environment.

Under the hood, the skill leverages Sleek’s REST API, requiring users to have a Sleek Pro plan account and an API key set in the environment (SLEEK_API_KEY). The skill’s code abstracts API interactions into commands the AI agent can invoke, effectively extending the agent’s domain knowledge with Sleek-specific design operations.

The stack is minimal from the user’s perspective: the skill is installed via Node.js’s npx CLI, and the AI agent interacts with Sleek remotely via the API. This decoupling keeps the skill lightweight while depending on Sleek’s cloud infrastructure.

What makes agent-skills technically interesting

What distinguishes this repo is its embodiment of the agent skills pattern as a concrete example in the AI tooling space. The pattern itself is gaining traction: AI assistants can now become platforms with installable capabilities rather than fixed-function tools.

The code quality appears straightforward and modular, focusing on wrapping Sleek’s API into skill commands. This separation makes it easy to update or add more skills without entangling core logic. The use of the npx skills CLI aligns with a developer-friendly DX, enabling interactive browsing and selective installation of skills.

The tradeoff is clear: the skill is tightly coupled to Sleek’s platform and requires a paid Pro plan. This limits accessibility and means the skill’s usefulness hinges on the API’s capabilities and stability. There is also only one skill currently, so the ecosystem is nascent compared to broader AI skill catalogs.

From a developer perspective, this repo is a practical demonstration of how to package API-driven capabilities as AI agent skills. It shows a clean pattern for extending AI agents with domain-specific operations via simple CLI tooling and environment-based authentication.

Quick start with agent-skills

Installation is straightforward using the npx skills CLI tool, which is designed for managing AI agent skills interactively or directly.

npx skills add sleekdotdesign/agent-skills

Or install the specific skill directly:

npx skills add sleekdotdesign/agent-skills -s sleek-design-mobile-apps

Requirements:

  • A Sleek account on the Pro plan or higher
  • An API key created at sleek.design/dashboard/api-keys, stored in the SLEEK_API_KEY environment variable

This minimal setup reflects the repo’s focus on ease of adding AI capabilities while relying on external platform services.

Where agent-skills fits in your toolkit

This repo is relevant if you’re exploring ways to extend AI assistants with targeted domain skills, especially in design automation. It demonstrates a clean pattern for integrating API-driven tools into AI workflows.

That said, the scope is narrow: it’s a single skill tied to a commercial design platform. If you’re looking for a broader AI skill ecosystem or open-source design tools, this might feel limiting.

The reliance on Sleek’s Pro plan and API keys also means this is not a fully open or standalone tool — the API is the bottleneck and point of dependency.

Still, as an example of the agent skills concept in action, it’s worth understanding. It shows how AI assistants can become modular platforms, installing skills like npm packages, which is a pattern likely to grow in AI developer tooling.

If your projects involve AI-based design workflows or you’re building AI agents that need extendable capabilities, this repo offers a useful case study and a starting point for building your own skills.

In production, expect to need to handle API rate limits, authentication renewals, and platform updates, as these will affect the skill’s reliability and feature set.

Overall, agent-skills by sleekdotdesign is a niche but practical illustration of modular AI agent extensibility centered on a real-world design API.

→ GitHub Repo: sleekdotdesign/agent-skills ⭐ 394