메뉴
Internal Python architecture rules loaded only when an explicitly invoked repository prompt directs the agent to this file.
기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.
설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/microsoft/FSQ --skill python-architecture명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
Internal project modification rules loaded only by the explicit /spec-driven prompt with a confirmed design path or direct project change request.
Internal diff-based project SPEC audit rules loaded only when an explicitly invoked repository prompt directs the agent to this file.
Optional internal design rules loaded only by the explicit /requirements-to-design prompt. Produces a confirmed design document for any requested change without changing SPEC or implementation files.
SKILL.md 표시 중
SOC 직업 분류 기준
| name | python-architecture |
| description | Internal Python architecture rules loaded only when an explicitly invoked repository prompt directs the agent to this file. |
| user-invocable | false |
| disable-model-invocation | true |
This is the internal Python architecture rules layer for an explicitly invoked SDD workflow. It helps choose the simplest adequate architecture, write Python-aware SPEC.md sections, implement against those specs, and audit code/spec synchronization.
Load this skill only when an explicitly invoked repository prompt directs the agent to this file. Ordinary discussion, explanation, review, planning, natural-language edit requests, skill-name mentions, and prose approvals must not trigger it.
Start simple and escalate only with evidence. Architecture exists to clarify ownership, isolate change, and protect business rules; it is not a template to apply everywhere.
Use for Python work involving:
Do not use for tiny one-off scripts unless the script is becoming a maintained module.
Choose one level and record the rationale in SPEC for non-trivial work:
| Level | Use when | Avoid when |
|---|---|---|
| 1. Script | One-off utility, no public API, little reuse | It has tests, users, or recurring changes |
| 2. Simple package | Small reusable code with clear functions/classes | Business rules span multiple workflows |
| 3. Layered application | API/CLI/worker with orchestration and persistence | Layers only pass data through |
| 4. Clean Architecture | Business rules must outlive framework/database choices | CRUD is simple and stable |
| 5. Light DDD | Rich domain behavior, invariants, bounded contexts | Domain is mostly data entry/retrieval |
Prefer the lowest level that keeps the module understandable and testable.
For Python modules, include this section in module SPEC.md:
## Python Architecture
- Architecture level: {1-5 and name}
- Public API: {exports, commands, endpoints, events, or classes}
- Internal modules: {files not imported outside this module}
- Domain boundaries: {where business rules live}
- Boundary models: {Pydantic schemas, DTOs, ORM models, serializers}
- Dependency direction: {allowed imports and forbidden imports}
- Rationale: {why this level is sufficient}
See references/module-spec-template.md for a fuller template.
__init__.py or the project-designated entry point._ when the project has no other convention.Before completion, check:
__init__.py exports, CLI commands, endpoints, or documented public symbols._internal files.Use references/audit-checklist.md for Python domain checks within the consolidated project implementation audit. Do not start a separate synchronization scan or repair loop.
references/architecture-levels.md: detailed level selection guidance.references/module-spec-template.md: Python module SPEC template.references/implementation-rules.md: implementation patterns and anti-patterns.references/audit-checklist.md: SPEC/code audit checklist.