ios-route-planner
Turn an iOS app idea or feature request into an Apple-native route before code is written — framework and symbol choices, state/data boundary, native surface, permission/entitlement/privacy matrix, fallback states, and a proportional verification plan. Use when starting a new iOS/iPadOS/watchOS/CarPlay/App Clip/visionOS app or feature, choosing between Apple frameworks, or untangling framework, permission, or system-surface confusion in an existing project.
- License 许可
- MIT
- 最近更新
- 2026-11-10
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- ios-route-planner
- description
- Turn an iOS app idea or feature request into an Apple-native route before code is written — framework and symbol choices, state/data boundary, native surface, permission/entitlement/privacy matrix, fallback states, and a proportional verification plan. Use when starting a new iOS/iPadOS/watchOS/CarPlay/App Clip/visionOS app or feature, choosing between Apple frameworks, or untangling framework, permission, or system-surface confusion in an existing project.
- allowed-tools
- Read, Grep, Glob, Write
- last_verified
- 1787097600000
02 资源文件 1
- 参考文档references/details.md6.7 KB
03 SKILL.md 正文
iOS Route Planner
Overview
Turn an idea into a route record and evidence plan before framework choices harden into implementation. This skill plans; the domain skills build. Hand off once the route is chosen:
| Lane | Skill |
|---|---|
| Targets, modules, extensions, schemes | ios-project-target-architect |
| SwiftUI screens and Liquid Glass | ios-native-design |
| Foundation Models, Vision, Core ML, Speech, Translation | ios-on-device-intelligence |
| Camera, audio, media, NFC, MusicKit, ShazamKit | ios-media-and-inputs |
| Persistence, sync, HealthKit, accessories | ios-data-and-device-services |
| StoreKit, Apple Pay, sign-in, Keychain, crypto | ios-commerce-identity-and-security |
| Widgets, Live Activities, extensions, background | ios-system-surfaces-and-background |
| Watch, CarPlay, App Clips, calling, push | ios-companion-communications |
| AR, RealityKit, Metal, SpriteKit, Game Center | ios-spatial-graphics-and-games |
| Tests, device proof, privacy manifest, TestFlight | ios-release-proof |
outcome -> narrow Apple route -> state/data boundary -> native surface -> permission/entitlement -> lifecycle/fallback -> verification evidence
Read before acting
- Inspect the actual repository and target: Xcode project/workspace, schemes,
deployment target, platforms/device families, modules, persistence, networking,
extensions, entitlements,
Info.plist, privacy manifest, existing system surfaces, and current tests. If no target exists, label the work bootstrap. - Start from the knowledge-base index, the framework selection questionnaire, and the idea-to-route recipe.
- The capability-first SDK atlas, framework catalog, and availability and device matrix are large. Grep them for the framework or capability name; do not read them whole.
- Before relying on a symbol, availability condition, entitlement, or system behavior, re-open the official page listed in the matching source registry file. The knowledge base was last checked against iOS 26 / Xcode 26 in August 2026.
Route workflow
- Intake. State the outcome in one sentence. Record entry point, primary action, accepted result, consequence of error, offline requirement, privacy sensitivity, supported platforms, and non-goals.
- Classify the capability: present/edit, persist/sync, capture/analyze, communicate, locate/map/weather, use protected data, control a device, transact/authenticate, expose to the system, share/export, run in the background, spatial/graphics/game, or companion surface.
- Select the narrowest Apple route that owns the capability. Prefer system-owned surfaces and first-party frameworks before a custom service. Name concrete symbols and rejected alternatives, and list which signatures still need an Xcode check.
- Draw the handoff:
input -> framework operation -> normalized evidence -> deterministic validation -> domain truth -> derived presentation -> system/companion handoff. Keep UI state, persistence, model output, and external system state distinct. - Build the state matrix before the happy path: ready, denied, restricted, unsupported, unavailable, loading, partial, stale, interrupted, cancelled, empty, invalid, conflict, completed — where relevant. Keep a manual or local-first path when it satisfies the goal.
- List configuration: permissions, usage descriptions, entitlements,
background modes, App Groups, associated domains, accounts/services, language
or model assets, hardware, server dependencies. Mark each
to-verify; never infer it from a framework name. - Choose proportional evidence: source, compile, fixture, simulator, physical device, two-device/accessory/vehicle, system surface, signed artifact, TestFlight/App Store, production. Name what each proves and what it cannot.
- Record the route in the project's planning docs or the design brief template, then stop unless the user asked for the build. If they did, hand the first narrow slice to the owning lane skill.
Fast path
For one feature, compare only the plausible lanes against outcome ownership,
target availability, setup cost, fallback, and required proof. Pick the route
with the least irreversible configuration; record the rest as rejected or
deferred. Freeze outcome -> trusted input -> smallest transformation -> user-visible destination and plan one slice with one source, one failure path,
and one proof task.
Route record
| Field | Required content |
|---|---|
| Outcome and consequence | User task, primary action, acceptable failure, reversibility |
| Selected route | Frameworks, symbols, platforms, rejected alternatives, open API questions |
| Data boundary | Source, normalization, persistence, sync, retention, deletion |
| UI/system boundary | Screen, extension, Watch/CarPlay/spatial surface, deep link, fallback |
| State/lifecycle | Permission, availability, cancel/interrupt/background/account states |
| Trust boundary | Validation, authorization, confirmation, idempotency, side effects |
| Configuration | Entitlements, usage strings, privacy manifest, modes, domains, servers |
| Tests | Fixtures, unit/UI/accessibility/performance, device and system runs |
| Evidence gaps | Claims not yet proven and the next smallest verification |
Change boundary
May inspect the project and write scoped planning, route, permission, and verification documents. Do not infer authorization for new accounts, cloud storage, analytics, paid services, background execution, health data, production credentials, deployment, or contacting users.
The capability decision table and refusal list are in references/details.md.
04 同仓库更多 Skills
no-description
暂无简介
good-skill
Fixture skill that satisfies every rule.
broken-link
Links to a reference file that does not exist.
Bad_Name
Name violates the kebab-case rule and does not match the directory.
skill-authoring
Write, review, and validate portable agent skills in SKILL.md format. Use when creating a new skill, editing an existing one, or asked whether a skill follows the Agent Skills spec.
weigh-bags
Record scale readings for each bag — empty (tare), outbound, current, final — and reconcile them against the itemised inventory to find unaccounted weight. Use when the user gives a bag weight, says how heavy something is, weighs their luggage, or asks how close to the limit they are. Appends to bags.yaml and runs scripts/reconcile.py.