Free SKILL.md scraped from GitHub. Clone the repo or copy the file directly into your Claude Code skills directory.
npx versuz@latest install linyihong-ai-skill-skills-app-development-guidancegit clone https://github.com/linyihong/Ai-skill.gitcp Ai-skill/SKILL.MD ~/.claude/skills/linyihong-ai-skill-skills-app-development-guidance/SKILL.md--- name: app-development-guidance description: Convert authorized app, API, embedded, firmware, and hardware-product findings into practical development guidance. Use when reverse-engineering lessons, product briefs, or hardware/firmware specs should become buildable app/API/SDK/firmware patterns, contracts, BDD/TDD plans, validation tests, release checks, and review checklists across mobile, web, backend, embedded, and future product types. --- # App Development Guidance Use this skill when APK analysis, app/API review, embedded/firmware review, or product development work reveals a reusable lesson that can improve future apps or hardware-backed products. The goal is to translate observed behavior, attack paths, failure modes, implementation tricks, hardware constraints, and protocol contracts into practical development requirements, buildable patterns, checklists, and validation tests. **Shared policy:** read [`shared-rules` index](../../shared-rules/README.md), [`dependency-reading.md`](../../shared-rules/dependency-reading.md), [`feedback-lessons.md`](../../shared-rules/feedback-lessons.md), [`reusable-guidance-boundary.md`](../../shared-rules/reusable-guidance-boundary.md), apply [`neutral-language.md`](../../shared-rules/neutral-language.md) when naming or summarizing docs, and apply [`goal-action-validation.md`](../../shared-rules/goal-action-validation.md) so important conclusions include a goal, action, validation, or reference source. If this skill or a related rule/template/lesson has changed, or the user asks to reload it, create a dependency read ledger (required files, files read, missing files marked `not applicable`, blocked items, validation), read the dependent docs before concluding, and close the Ai-skill writeback transaction before returning to project work. Lessons in `feedback_history/` should reference those files, not duplicate shared rules. **Cross-skill references:** follow [`cross-skill-references.md`](../../shared-rules/cross-skill-references.md). When another skill hands off sanitized analysis artifacts, this skill consumes the development-relevant contract, risk, validation, and open-question details without copying the source skill's analysis workflow. ## When To Use - Turning APK analysis findings into practical development guidance. - Turning observed app/API techniques into buildable patterns for your own app. - Turning APK analysis documents into app-related tools, SDKs, clients, mocks, fixture-driven implementations, contract tests, or rebuilt features. - Turning product briefs into BDD, Domain Model Contracts, API Contracts, implementation slices, and tests. - Validating product briefs before using them as implementation input: goals, users, scope, non-goals, assumptions, success criteria, constraints, dependencies, risks, and blocker questions. - Turning hardware, sensor, protocol, firmware, or embedded product specs into datasheet/protocol contracts, hardware context contracts, firmware BDD, driver/service/application boundaries, host/target tests, and bring-up validation. - Backfilling implemented-first projects with document precedence, product/rule traceability, BDD-to-test closure, minimum doc-sync matrices, generated-client checks, and explicit canceled/deferred/out-of-scope decisions. - Turning OpenAPI/schema contracts into typed clients, SDKs, mocks, fixtures, and provider/consumer tests without hand-copied endpoints. - Turning vendor or third-party API integrations into sanitized integration excerpts, fixture tests, live-test gates, webhook/idempotency checks, and secret-safe docs. - Turning IDE extensions, CLIs, linters, static analyzers, code generators, and internal tools into rule catalogs, pure kernels, adapters, diagnostics/commands, fixtures, and integration tests. - Classifying app changes as new requirements, bug fixes, refactors, or hardening work before code; new requirements must update planning docs before implementation. - Separating legacy regression protection from new-code validation, including BDD/TDD, changed-code coverage, mutation tests, property-based tests, contract tests, database regression tests, and human review. - Adding performance testing to development and release gates when changes can affect latency, throughput, resource usage, startup, background work, database access, batching, caching, concurrency, or external-call volume. - Backfilling missing development documents for existing, already implemented projects; Product Brief gaps can be marked unknown, but BDD behavior must be completed from observed implementation evidence. - Reviewing app/API design for replay resistance, token safety, transport security, local storage, logging, and release hardening. - Creating PR/release checklists for mobile, web, backend/API, and future app types. - Deciding what should be validated by tests, fixtures, runtime checks, or server-side controls. ## Out Of Scope - Breaking into apps or services without authorization. - Storing target-specific hosts, secrets, endpoints, tokens, real user data, or private business conclusions in this reusable skill. - Treating client-side hardening as a replacement for server-side authorization, rate limits, fraud controls, or audit logs. ## Quick Start(Routing) See [`runtime/onboarding/app-development-guidance-quickstart.md`](../../runtime/onboarding/app-development-guidance-quickstart.md) for the full 15-step quick start. Routing summary: 1. Identify source → 2. Validate planning docs → 3. Classify request → 4. New req: planning docs first → 5. Bug fix: before/after code rules → 6. Test strategy before code → 7. Product brief → `process/` → 8. Existing project: backfill → 9. Blocker questions → 10. Risk statement → 11. Choose control layer → 12. Validation method → 13. Classify guidance → 14. Linked updates → 15. Immature lesson → `feedback_history/`. ## Default Workflow ### 新分層路徑(優先讀取) | 用途 | 路徑 | |------|------| | 開發流程(Product Brief → BDD → 實作 → 驗證) | [`workflow/app-development-guidance/development-process.md`](../../workflow/app-development-guidance/development-process.md) | | 執行流程(Change Intake → Docs-First BDD → SDK Defect → Performance Gate) | [`workflow/app-development-guidance/execution-flow.md`](../../workflow/app-development-guidance/execution-flow.md) | | 分析方法(Risk Translation、Controls Catalog、Platforms/Languages Catalog) | [`analysis/app-development-guidance/README.md`](../../analysis/app-development-guidance/README.md) | | 工程智慧(Heuristics、Anti-patterns) | [`intelligence/engineering/app-development-guidance/README.md`](../../intelligence/engineering/app-development-guidance/README.md) | | 產出格式與品質門檻 | [`workflow/app-development-guidance/artifact-gates.md`](../../workflow/app-development-guidance/artifact-gates.md) | | Review Checklist | [`workflow/app-development-guidance/review-checklist.md`](../../workflow/app-development-guidance/review-checklist.md) | ### 舊路徑(保留向後相容) | 用途 | 路徑 | |------|------| | WORKFLOW.md(舊執行流程) | [`WORKFLOW.md`](WORKFLOW.md) | | process/(舊開發流程) | [`process/`](process/) | | CHECKLIST.md + checklists/ | [`CHECKLIST.md`](CHECKLIST.md)、[`checklists/`](checklists/) | | DOCUMENTATION.md(舊產出格式) | [`DOCUMENTATION.md`](DOCUMENTATION.md) | | controls/ + platforms/ + languages/ + implementation/ | [`controls/`](controls/)、[`platforms/`](platforms/)、[`languages/`](languages/)、[`implementation/`](implementation/) | ## Output Style & Artifact Gates See [`workflow/app-development-guidance/artifact-gates.md`](../../workflow/app-development-guidance/artifact-gates.md) for output format and quality gates. ## Feedback Loop See [`shared-rules/feedback-lessons.md`](../../shared-rules/feedback-lessons.md) for the feedback lesson template and workflow. See [`feedback/`](../../feedback/) for the feedback promotion pipeline. **Cross-skill link:** if the lesson came from APK analysis, keep analysis mechanics in [`apk-analysis`](../apk-analysis/) and put development guidance, implementation patterns, and validation checklists here. When consuming a Feature Reconstruction Handoff from [`apk-analysis`](../apk-analysis/), this skill owns the development conversion: BDD, Domain Model Contract, Architecture Contract, API / Interface Contract, Error Handling Contract, implementation slices, fixtures, tests, and review checklists. If the requested output is an app-related tool, SDK, client, mock, fixture-driven implementation, contract test, or rebuilt feature, apply this skill before implementation planning and surface blocker questions. Keep raw APK evidence, target-specific hosts, tokens, accounts, and private product conclusions in the project analysis docs.