Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
Version 0.3.1 - Added dedicated reference guides for Clean Architecture, MVI, MVVM, Reactive, TCA, VIPER, and an architecture selection guide. - Improved handling for explicit architecture requests: perform a fit check, explain mismatches if any, and suggest the best-fit alternative. - For mismatch cases, proceed with a risk-mitigated plan if the user insists, highlighting risks up front. - Added output requirements to include fit results and rationale when an architecture is specified. - Clarified when to reference selection guides and how to justify architecture recommendations. - No changes to core deliverable structure or technical guidance flow.
What This Skill Does
Guides architecture selection and implementation for Swift codebases across MVVM, MVI, TCA, Clean Architecture, VIPER, and Reactive patterns. Covers new features, refactors, PR reviews, and debugging for both SwiftUI and UIKit projects. Validates fit before committing to an architecture and flags mismatches with alternatives.
Rather than applying a single pattern universally, it runs a fit check against your actual constraints, UI stack, and existing conventions before recommending or proceeding.
When to use it
- Choosing between TCA and MVVM for a new SwiftUI screen
- Refactoring a UIKit module to Clean Architecture incrementally
- Reviewing a pull request for VIPER layer violations
- Debugging state management issues in a Combine-heavy codebase
- Designing dependency injection boundaries for a multi-screen flow
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: I'm adding a multi-step onboarding flow in SwiftUI. What architecture should I use and how do I structure it?
- 1Captures task type (new feature), UI stack (SwiftUI), and scope (multi-screen flow)
- 2Loads selection-guide.md and evaluates state complexity, effect orchestration, and team familiarity
- 3Recommends TCA with a brief explanation of why it fits multi-screen state coordination
- 4Produces directory layout, state/dependency types, async cancellation strategy, and test stubs scoped to onboarding
- 5Outputs the TCA PR review checklist adapted to the onboarding feature
A concrete file structure, typed state and action definitions, an async effect plan with cancellation, and a feature-specific review checklist