AI product Open source
BAMA is a set of agent skills for researching, cloning, designing, building, optimizing, and testing native macOS applications with Swift and SwiftUI. Its workflow has the agent open and screenshot a reference app, create a screen index, turn the screens into user stories with checkable acceptance criteria grouped into milestones, and implement one story per commit. Before commits it snapshots pages in light and dark mode and runs tests. Skills cover app bootstrapping, market and design research, computer-use-based cloning, Swift and SwiftUI development, macOS integrations, performance, architecture, design systems, layout checks, app icons, building, and testing; an optional command-line and MCP server can let agents or scripts drive the app. The repository also includes build plans for small Mac utilities. It is distributed as installable skills for Claude Code, Codex, Cursor, Gemini CLI, and other compatible agents, and is licensed under the MIT License.
1 use taken from transcripts — each links to the moment in the video.
A set of agent skills that let a coding agent clone any app as a native macOS app in Swift and SwiftUI, from screenshotting a reference app to building user stories one commit at a time.
1 in the library.