MCP Context Template is a repository designed to centrally manage AI tools (MCP) to reduce context sharing costs and improve development productivity.
It serves as a central context repository that helps modern AI tools like Claude Code and Cursor AI consistently understand your company's codebase and development practices.
- Common context repository managed independently from individual projects
- Providing reference document structure to improve collaboration efficiency with AI tools
- Containing examples and guides that can be used by different teams and organizational layers
- Providing consistent AI usage experience across projects
- Reducing the burden of duplicate context management
ai-context/
βββ README.md
βββ context/
β βββ backend/
β β βββ domain-overview.md
β β βββ service-patterns.md
β β βββ test-guidelines.md
β βββ frontend/
β β βββ component-structure.md
β β βββ naming-convention.md
β βββ mobile/
β β βββ network-layer.md
β βββ shared/
β βββ business-terms.md
β βββ naming-style.md
βββ prompts/
β βββ refactoring.md
β βββ test-generation.md
β βββ code-review.md
βββ examples/
βββ cursor-snippets.md
βββ claude-usage.md
- Set
ai-context/
as a context path in your project's.cloderc
or Cursor settings - Explicitly reference guides in Claude Code prompts with "Please refer to the following guide: [path]"
- Example: Request refactoring based on
context/backend/service-patterns.md
Refactoring request:
Please refactor the following code according to the service patterns defined in context/backend/service-patterns.md.
[INSERT CODE]
- Contribute via Pull Request
- Write all convention documents in Markdown
- Propose significant context structure changes as issues first for discussion
- Write context concisely and clearly for easy AI understanding
Example from prompts/test-generation.md
:
## Purpose
Claude prompt example for automating service unit test generation
## Prompt
Please write unit test code for the following service class method:
- Test framework: JUnit5
- Mocking: MockK
- Use Given-When-Then format
- Target method:
```kotlin
fun getUserById(id: Long): User
- New developer onboarding - quickly understand code styles and patterns
- Automation of repetitive coding tasks - generating test code, boilerplate
- Code review automation - generate review comments with consistent standards
- Documentation support - assist in codebase documentation tasks
MCP Context Templateμ AI λꡬ(MCP)λ₯Ό "μ€μ μ§μ€μ"μΌλ‘ κ΄λ¦¬νμ¬ μ»¨ν μ€νΈ 곡μ λΉμ©μ μ€μ΄κ³ κ°λ° μμ°μ±μ λμ΄κΈ° μν λ ν¬μ§ν 리μ λλ€.
Claude Code, Cursor AI λ± μ΅μ AI λꡬλ€μ΄ μΌκ΄λ λ°©μμΌλ‘ νμ¬μ μ½λλ² μ΄μ€μ κ°λ° κ΄νμ μ΄ν΄ν μ μλλ‘ λλ μ€μ 컨ν μ€νΈ μ μ₯μλ‘ κΈ°λ₯ν©λλ€.
- νλ‘μ νΈμ λ 립μ μΌλ‘ κ΄λ¦¬λλ κ³΅ν΅ μ»¨ν μ€νΈ μ μ₯μ
- AI λꡬμμ νμ ν¨μ¨μ λμ΄κΈ° μν μ°Έκ³ λ¬Έμ ꡬ쑰 μ 곡
- νλ³, κ³μΈ΅λ³λ‘ νμ© κ°λ₯ν μμμ κ°μ΄λ μλ‘
- νλ‘μ νΈ κ° μΌκ΄λ AI μ¬μ© κ²½ν μ 곡
- μ€λ³΅ 컨ν μ€νΈ κ΄λ¦¬ λΆλ΄ κ°μ
ai-context/
βββ README.md
βββ context/
β βββ backend/
β β βββ domain-overview.md
β β βββ service-patterns.md
β β βββ test-guidelines.md
β βββ frontend/
β β βββ component-structure.md
β β βββ naming-convention.md
β βββ mobile/
β β βββ network-layer.md
β βββ shared/
β βββ business-terms.md
β βββ naming-style.md
βββ prompts/
β βββ refactoring.md
β βββ test-generation.md
β βββ code-review.md
βββ examples/
βββ cursor-snippets.md
βββ claude-usage.md
- νλ‘μ νΈ λ΄
.cloderc
λλ Cursor μ€μ μμai-context/
λ₯Ό context pathλ‘ μ§μ - Claude Code ν둬ννΈ μμ± μ "λ€μ κ°μ΄λλ₯Ό μ°Έκ³ ν΄μ£ΌμΈμ: [κ²½λ‘]"μ κ°μ΄ λͺ μμ μΌλ‘ μΈκΈ
- μμ:
context/backend/service-patterns.md
μ κΈ°λ°ν 리ν©ν λ§ μμ²
리ν©ν λ§ μμ²:
context/backend/service-patterns.mdμ μ μλ μλΉμ€ ν¨ν΄μ λ°λΌ λ€μ μ½λλ₯Ό 리ν©ν λ§ν΄ μ£ΌμΈμ.
[μ½λ μ½μ
]
- Pull Requestλ₯Ό ν΅ν΄ κΈ°μ¬
- λͺ¨λ 컨벀μ λ¬Έμλ MarkdownμΌλ‘ μμ±
- ν° μ»¨ν μ€νΈ ꡬ쑰 λ³κ²½μ μ΄μλ‘ λ¨Όμ μ μ ν λ Όμ
- 컨ν μ€νΈλ κ°κ²°νκ³ λͺ ννκ² μμ±νμ¬ AIκ° μ΄ν΄νκΈ° μ½κ² ꡬμ±
prompts/test-generation.md
μμ:
## λͺ©μ
μλΉμ€ λ¨μ ν
μ€νΈ μλνλ₯Ό μν Claudeμ© ν둬ννΈ μμ
## ν둬ννΈ
μλΉμ€ ν΄λμ€μ λ©μλμ λν΄ λ¨μ ν
μ€νΈ μ½λλ₯Ό μμ±ν΄μ£ΌμΈμ.
- ν
μ€νΈ νλ μμν¬: JUnit5
- λͺ¨νΉ: MockK
- Given-When-Then νμ μ¬μ©
- ν
μ€νΈ λμ λ©μλ:
```kotlin
fun getUserById(id: Long): User
- μλ‘μ΄ κ°λ°μ μ¨λ³΄λ© - μ½λ μ€νμΌκ³Ό ν¨ν΄μ λΉ λ₯΄κ² μ΄ν΄
- λ°λ³΅μ μΈ μ½λ μμ μλν - ν μ€νΈ μ½λ, 보μΌλ¬νλ μ΄νΈ μμ±
- μ½λ 리뷰 μλν - μΌκ΄λ κΈ°μ€μΌλ‘ 리뷰 μ½λ©νΈ μμ±
- λ¬Έμν μ§μ - μ½λλ² μ΄μ€ λ¬Έμν μμ 보쑰