An Apple dev expert that fits inside your AI assistant

What it does

Axiom is a knowledge layer and tooling suite that turns generic AI coding assistants into Apple-platform specialists. It ships 260+ skills covering Swift 6, SwiftUI, and concurrency, plus 41 autonomous agents that scan for memory leaks and data races. Four native Go utilities—xclog, xcsym, xcui, and xcprof—let an AI capture simulator logs, symbolicate crashes, drive UI via accessibility IDs, and record CPU traces without ever opening Instruments.

The interesting bit

Most prompt libraries are just text files; Axiom TDD-tests every skill against “real developer pressure scenarios” and bundles actual binaries to bridge the gap between an LLM and Xcode’s opaque toolchain. It is effectively a senior iOS engineer compressed into a plugin architecture.

Key highlights

• 260+ skills spanning UI, data, networking, accessibility, and performance
• 41 autonomous agents audit for memory leaks, concurrency violations, and build failures
• Native Go tools for console capture, crash symbolication, simulator UI automation, and xctrace profiling
• Skills tested against realistic failure scenarios rather than toy examples
• Tracks Apple OS 27 betas continuously while keeping OS 26 guidance stable

Caveats

• The project is currently on a 27.x beta cycle tracking Apple’s OS 27 developer betas, so bleeding-edge coverage is still landing
• README stats (skill/agent counts) are auto-maintained by scripts, which suggests the surface area changes frequently

Verdict

Worth a look if you use Claude Code, Codex, or Cursor for iOS/macOS work and are tired of explaining Xcodebuild zombies to an AI. Probably overkill if you only dabble in Apple platforms occasionally.