Configures Claude Code hooks and Codex hooks.json/notify. Use when adding PreToolUse guards, Stop hooks, managed hooks, format-on-save, preflight, audits, or worktree/budget hooks.
日本語の概要は準備中です。原文の説明を表示しています。
Designs and audits native iOS interfaces. Use when reviewing or refining SwiftUI layout, typography, Liquid Glass, navigation, or dashboards on a freshly verified build.
インストールする前に、エージェントに与えられる指示の中身を確認できます。
Use this skill for visual design decisions and design-focused audits in any native iOS app. Prefer it when the user needs HIG-aligned screen structure, current Apple-native visual defaults, or a screenshot-to-fix loop after a fresh verified build/install/launch. If the request is really about greenfield scaffolding, CLI build loops, scheme selection, or broader iOS implementation workflow, route to software-ios-native and return here once the question becomes visual structure or design quality.
.contentShape() when the visible glyph is smaller..sheet + .presentationDetents) for detail over inline expansion.ScrollView > LazyVStack with .scrollTransition for narrative screens; reserve List(.insetGrouped) for settings-style hubs..frame(maxWidth: .infinity, alignment: .leading)) so short rows do not shrink beside longer rows.PlainButtonStyle can provide state feedback; its presence alone does not prove a dead tap.iOS 17+ polish: .scrollTransition, .sensoryFeedback, .symbolEffect, .contentTransition(.numericText). Standardize scroll reveals into a shared ViewModifier (e.g., .appScrollReveal()).scaleEffect(0.94) with 100 ms ease-out is a starting token, not an Apple requirement. Retain native feedback when it already conveys the action and respect Reduce Motion..presentationDetents([.medium, .large]) on sheets. Use .presentationBackgroundInteraction(.enabled(upThrough:)) when the surface behind the sheet should stay interactive at peek height.ZStack { background; visualization; controls; .sheet(persistent) } — no card wrappers, no scroll.Picker(.segmented) + Menu overflow over custom material-backed button strips.TabView for 3–5 peer top-level sections on iPhone; verify fit and task structure. Reach for NavigationSplitView (sidebar-adaptive) when (a) the app is iPad/Mac-first and needs a persistent list-detail relationship, (b) there are 6+ top-level destinations that would force a tab-bar "More" overflow, or (c) the same information architecture must scale from iPhone compact width to iPad/Mac without a redesign. NavigationSplitView collapses to a stack on iPhone automatically — it is the correct default for adaptive multi-platform apps, not just a Mac/iPad nicety. Do not default to a custom hamburger-drawer sidebar on iPhone; it is not a system pattern and fails discoverability audits.This skill covers the visual audit only. Strings, plurals, catalog parity and the translation pipeline are owned by software-localisation.
.accessibilityLabel. Combine multi-element rows with .accessibilityElement(children: .combine).set -o pipefail; xcodebuild ... 2>&1 | xcbeautify) plus simctl; preserve the build exit status. Route packaging or simulator-health issues to the runtime-debugging skill. See references/xcodebuildmcp-design-loop.md for setup and commands.For every changed surface, select states by risk rather than presenting one ideal screenshot: populated, loading, empty, error, permission-denied, offline, and destructive-action states where applicable. Cross them with the smallest configuration set that can reveal the change's failure mode: smallest and largest supported width, AX Dynamic Type for layout changes, light/dark plus Increase Contrast or Reduce Transparency for material changes, and VoiceOver focus/announcement order for semantic changes. Record unreachable states and why.
Use the Feel Bar as a team review rubric, selecting applicable rows for the screen. The three-failure debt threshold is a team default, not a platform rule; decorative motion or haptics are not required to pass. Pass/fail rows are in references/ios-craft-and-feel.md → The Feel Bar.
Before writing or reviewing screen code, check these patterns from references/design-craft-patterns.md:
.contentTransition(.numericText); use gradient score bars. See references/ios-component-patterns.md.ButtonStyle), haptics (.sensoryFeedback), SF Symbol animation (.symbolEffect(.bounce)), expandable sections, entrance animations on sheets..accessibilityHint. Multi-element rows use .accessibilityElement(children: .combine). Navigation chevrons need explicit labels. See references/ios-accessibility-patterns.md.Apply the Runtime Proof Gate before reviewing. Compare before/after screenshots alongside Dynamic Type, appearance, and device-width evidence. If images cannot explain the failure, capture focused layout, appearance, or accessibility logs after launch; use references/ai-design-review.md for the review prompt and proof shape.
Widgets, Live Activities, Control Center custom controls, and Lock-screen surfaces are first-class iOS design surfaces. This skill's patterns apply (typography, Dynamic Type, materials, touch targets, accessibility) but the platform constraints differ — widget timelines, size families, interactive affordance rules. Load references/ios-surfaces-reference.md for widget rendering modes, ControlWidget templates, Live Activity presentations, and Icon Composer decisions. Look up target-OS and SDK availability at the surface before specifying it.
glassEffect, GlassEffectContainer, morphing, fallback for pre-iOS 26, accessibility interactionsOptional helpers for a proven-build design loop. If the project uses a Makefile, make can replace the individual scripts. If the app cannot build/install/launch cleanly, route to ../software-ios-runtime-debugging/SKILL.md.
.xcodebuildmcp/config.yamlFor the full anti-pattern catalog — SwiftUI state/identity, navigation, layout, color/Dark Mode, animations, controls, sheets, data display, Canvas, and runtime footguns — see references/swiftui-design-antipatterns.md.
| Check | Pass condition | Fail condition |
|---|---|---|
| Standard structure | Recommendation maps to a standard iOS pattern | Custom pattern used without justification |
| Screenshot origin | Freshly installed and launched build | Screenshot from a prior install or simulator session not tied to current build |
| Typography | Dynamic Type text styles throughout | Fixed-size design tokens |
| Color and materials | Contrast and hierarchy preserved in light and dark mode | Material or color breaks in one appearance |
| Touch targets | ≥ 44 × 44 pt; .contentShape() used where glyph is smaller | Target below minimum; dead taps on tappable rows |
| Notification surface | Verified on real iPhone in correct APNs environment | Simulator-only screenshots for notification UX |
| Dashboard heuristics | Labeled as team default, not platform rule | Presented as HIG requirement without citation |
| XcodeBuildMCP guidance | Verified against upstream docs | Repeated from memory or prior session |
| System surfaces | Widget rendering modes, controls, and icon appearances verified for target OS | Only a full-color in-app screenshot supplied |
At structure selection, read Apple Developer Releases and the target component's HIG page from data/sources.json. Design for the shipped target OS; separately test the minimum deployment target. At symbol, icon, or system-surface selection, use the lookup in its reference rather than inferring availability from a tool release. At runtime verification, check upstream XcodeBuildMCP command/configuration docs before invoking the wrappers.
When prior decisions or pitfalls are relevant, consult learnings.consolidated.md if present; use learnings.md only for needed history or as the available fallback. Otherwise skip both.
After applying it, if you encountered a pattern worth remembering, a mistake worth preventing, or a domain fact that surprised you, append one dated bullet to learnings.md via agents-skills-feedback-loop/scripts/append_learning.py. Do not modify SKILL.md itself.
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Configures Claude Code hooks and Codex hooks.json/notify. Use when adding PreToolUse guards, Stop hooks, managed hooks, format-on-save, preflight, audits, or worktree/budget hooks.
日本語の概要は準備中です。原文の説明を表示しています。
Configures and hardens Claude Code and Codex MCP servers. Use when connecting databases, APIs, SaaS, building servers, or serving a clearance-filtered knowledge base.
日本語の概要は準備中です。原文の説明を表示しています。
Owns instruction files: AGENTS.md, CLAUDE.md, personal and repo rules. Use when writing, pruning, auditing them, sharing rules across Claude and Codex, or fixing ignored rules.
日本語の概要は準備中です。原文の説明を表示しています。
Creates and audits agent skills: SKILL.md, references, scripts, runtime metadata. Use when writing, validating, or security-reviewing a skill, or fixing truncated skill listings.
日本語の概要は準備中です。原文の説明を表示しています。
Adds per-skill learnings loops for dated patterns, mistakes, and domain facts. Use when wiring skill memory, consolidation, or drift audits.
日本語の概要は準備中です。原文の説明を表示しています。
Chooses subagent, team, workflow, or debate and launches it on Claude Code or Codex. Use when delegating, running agent review boards, or installing shared agents.
日本語の概要は準備中です。原文の説明を表示しています。