architecture
/architecture
Pomaga projektować architekturę systemów, oceniać kompromisy, granice usług, schematy baz i kontrakty API oraz zapisywać ADR.
- Autor
- hmfareed ↗
- Repozytorium
- Frontier-Engineer ↗
- Licencja
- MIT
- Ostatnia zmiana repozytorium
- 13.11.2026
Ocena i bezpieczeństwo
- Pliki w paczce
- 2
- Skrypty
- 0
- Komentarze HTML
- 0
- Adresy URL w SKILL.md
- 0
Brak sygnałów. Repozytorium zmienione 13.11.2026, bez odwołań do wycofanych modeli i API.
MIT, otwarta: można używać i zmieniać, z zachowaniem informacji o autorze.
Co robi i kiedy
Pomaga projektować architekturę systemów, oceniać kompromisy, granice usług, schematy baz i kontrakty API oraz zapisywać ADR.
Oryginalny opis (EN)
Use when designing system architectures, evaluating architectural trade-offs, planning service boundaries, database schemas, API contracts, or recording Architecture Decision Records (ADRs).
Jak działa
- Purpose
- Activation criteria
- Non-activation guidelines
- Required context and inputs
- Operating procedure
- Technical and engineering standards
- Verification and evidence requirements
- Expected outputs and deliverables
- Risks, boundaries and failure modes
- Progressive disclosure references
Podgląd SKILL.md
--- name: "architecture" description: "Use when designing system architectures, evaluating architectural trade-offs, planning service boundaries, database schemas, API contracts, or recording Architecture Decision Records (ADRs)." --- # Architecture — System Architecture and Technical Design > Generated from `specs/skills/architecture.md` by `scripts/generate-antigravity-skills.ts`. Edit the canonical specification, not this file. ## Purpose The Architecture skill provides a disciplined, evidence-based procedure for designing reliable, scalable, and maintainable software systems. It ensures that system decomposition, interface boundaries, persistence layers, and communication topologies are chosen through explicit trade-off evaluation rather than trend-following or premature complexity. ## Activation criteria Activate this skill when: - Designing a new system, major subsystem, or greenfield application. - Evaluating architectural alternatives (e.g., monolith versus microservices, synchronous versus event-driven). - Defining cross-cutting service boundaries, API contracts, and shared domain models. - Planning structural database migrations, multi-tenant isolation, or caching hierarchies. - Documenting significant architectural decisions in Architecture Decision Records (ADRs). ## Non-activation guidelines Do not activate this skill when: - Resolving isolated defects within an existing component without changing public contracts. - Adjusting frontend visual styles, layouts, or component animations. - Writing unit tests for already defined interfaces. - Updating dependencies within compatible semver bounds. ## Required context and inputs Before initiating architectural design, ensure the following inputs are gathered and verified: 1. **Functional Requirements:** Core user journeys, input payloads, business invariants, and expected outputs. 2. **Non-Functional Requirements:** Throughput targets (requests/sec), latency bounds (p95/p99), availability SLAs, data volume projections, and retention policies. 3. **Repository Evidence:** Existing framework conventions, directory structures, language versions, existing database engines, and deployment manifests. 4. **Constraints:** Team size, operational budget, security/compliance regimes (e.g., HIPAA, GDPR, SOC2), and backwards-compatibility requirements. ## Operating procedure Follow this eleven-step engineering sequence: 1. **Understand Requirements:** Extract explicit requirements and identify unspoken operational constraints. 2. **Inspect Existing Repository:** Map existing component boundaries, entry points, data models, and dependency graphs. Never guess what libraries or databases exist. 3. **Identify System Boundaries:** Group related capabilities by bounded context, identifying clear ownership of state. 4. **Evaluate Architectural Alternatives:** Formulate at least two viable architectural approaches (e.g., modular monolith vs. microservices, synchronous REST vs. async messaging). 5. **Explain Trade-offs:** Score each alternative against operational complexity, performance, scalability, failure domains, and delivery timeline. 6. **Recommend Architecture:** Propose a single primary recommendation with explicit rationale justifying why rejected options were discarded. 7. **Define Components and Interfaces:** Detail service responsibilities, public interfaces, synchronous HTTP/gRPC contracts, and event schemas. 8. **Define Data Flows:** Trace write paths (commands) and read paths (queries), specifying ACID transactional boundaries and eventual consistency tolerances. 9. **Identify Reliability and Security Risks:** Model single points of failure (SPOFs), backpressure management, circuit breaking, rate limiting, and zero-trust authentication boundaries. 10. **Formulate Migration Plan:** If modifying existing systems, define incremental, reversible phases (e.g., Strangler Fig pattern, dual-writing). 11. **Document Decisions:** Record each major conclusion in an Architecture Decision Record (ADR) adhering to the standard format. ## Technical and engineering standards - **Simplicity First:** Prefer modular monoliths and local procedural calls over distributed microservices unless independent organizational scaling or hardware needs demand separation. - **Explicit Contracts:** All inter-boundary communication must be backed by typed schemas (JSON Schema, OpenAPI, Protocol Buffers, or TypeScript interfaces). - **Idempotency:** All asynchronous message handlers and non-idempotent HTTP mutation endpoints (`POST`) that touch financial or state-critical operations must specify idempotency key handling. - **Persistence Isolation:** Services must not share raw database tables. Communication across bounded contexts must traverse public APIs or verified events. - **Failure Isolation:** Any external API or database call must include configured timeouts, retries with exponential backoff and jitter, and fallback behavior. ## Verification and evidence requirements To verify an architectural proposal: - Validate schema contracts against actual consumer requirements. - Demonstrate that migration steps are backward-compatible and non-destructive. - Verify that no circular dependencies exist between identified modules. - Cite repository source files to prove compatibility with installed runtimes and dependencies. ## Expected outputs and deliverables - **Architecture Design Document:** Clear prose detailing component decomposition, data models, and network boundaries. - **Interface Specifications:** OpenAPI schemas, TypeScript types, or Protobuf definitions for new or modified contracts. - **Architecture Decision Record (ADR):** Formatted Markdown document recording context, options considered, final decision, and consequences. - **Migration & Rollback Plan:** Phased execution checklist with clear criteria for aborting or rolling back. ## Risks, boundaries and failure modes - **Over-Engineering Risk:** Recommending Kubernetes, Kafka, or microservices for small-scale applications adds operational drag without technical benefit. - **Under-Specified Contracts:** Leaving API field types or nullability vague leads to implementation drift and integration failure. - **Destructive Changes:** Proposing schema drops or breaking API changes without backward-compatibility deprecation windows is an architectural failure. ## Progressive disclosure references For detailed decision frameworks, trade-off matrices, and ADR authoring templates, refer to: - [Trade-off Matrix and ADR Authoring Guide](./references/trade-off-and-adr-guide.md)
Pliki w paczce (2)
- SKILL.md 6 KB
- references/trade-off-and-adr-guide.md 5 KB
Wersje
| Repozytorium | Wynik | Licencja |
|---|---|---|
| Ta strona · hmfareed/Frontier-Engineer | 90 | MIT |
| jikig-ai/soleur | 100 | Apache-2.0 |
| Come-In-Labs/agent-skills | 100 | MIT |
| dedujeong/gemba | 100 | Apache-2.0 |
| workbench-soft/agent-skills | 100 | MIT |
| boshu2/agentops | 100 | Apache-2.0 |
| Kanevry/session-orchestrator | 100 | MIT |
| mcouthon/agents | 100 | MIT |
| FrancoStino/opencode-skills-collection | 100 | MIT |
Migawka bazy z 13.11.2026 · wersja paczki 32d1dc7d8e96 (skrót SHA-256 całej paczki)
Pobrania w ostatnich 30 dniach: 0 · ranking
Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.
Podobne skille
- 95adr
Tworzy lub recenzuje Architecture Decision Records dla decyzji architektonicznych, trwałych kompromisów, nowych zależności, kontraktów i wzorców. - 100architecture-designer
Projektuje architekturę wysokiego poziomu, tworzy diagramy i ADR, ocenia kompromisy technologiczne, planuje skalowalność oraz przegląda istniejące projekty. - 100architecture-designer
Projektuje i przegląda architekturę systemów: diagramy, rekordy ADR, kompromisy technologiczne, skalowanie i wzorce infrastruktury. - 100backend-development
Pomaga projektować API, schematy baz i architekturę systemu backendowego, a także rozwijać go przez testy. - 100architecture-design
Projektuje architekturę systemu full-stack: diagramy C4, kontrakty API, diagramy sekwencji, topologie wdrożenia i ADR, przy projektowaniu nowych systemów, wyborze stosu i granic usług przed implementacją. - 100architecture-design
Projektuje lub ocenia architekturę systemu: granice usług, modele danych, kontrakty API, topologię infrastruktury, wybór bazy i analizę zależności, w tym monolit kontra mikroserwisy.