Back to directory

Appllama

A pair of agent skills that study the real screens of top-grossing mobile apps through an MCP research tool, then build native-quality Expo and React Native screens held to a simulator-verified motion and design bar.

Appllama

Appllama Skills give a coding agent a design library of top-grossing mobile apps' real screens, flows and UI patterns, along with revenue and download context, and turns that library into a working method for building new mobile screens. One skill, appllama-usage, is the research engine: it works with the Appllama MCP server like a design director, with a full tool map and playbooks for studying an app category from scratch, improving an existing screen, or researching a specific flow or UI element.

The second skill, appllama-app-design-skill, is the build bar: it targets native-feeling Expo and React Native screens held to Apple Human Interface Guidelines fidelity, semantic colors, native controls, navigation that behaves correctly for pushes, replaces, sheets and one-way doors, a strict motion bar for what should animate and how, generated image assets, and a full-motion simulator loop that scrubs whole recorded flows frame by frame rather than relying on static screenshots. This design skill can be installed and used on its own, without the Appllama MCP connected.

Both install with one command and work with Claude Code, Cursor, Codex and more than seventy other agents, and are released under the MIT license.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.