Skip to main content
Glama
deenesjakoruzh
k0ry

Shop MCP Server

by k0ry
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Local

Shop MCP Server

온라인 상점 SQLite 데이터베이스 분석을 위한 로컬 MCP 서버입니다. AI 에이전트가 MCP 도구를 호출하면 서버가 요청을 안전한 SQL SELECT로 변환하고, shop.db를 직접 읽어 구조화된 결과를 반환합니다.

Host-приложение
├── AI-агент / LLM
└── MCP-клиент
        │ JSON-RPC / stdio
        ▼
    Shop MCP Server
        │ read-only SQLite
        ▼
    data/shop.db

기능

  • HTTP 및 별도 DB 서버 없이 SQLite에 직접 연결;

  • MCP 전송 stdio;

  • tools/list를 통한 query_shop 및 inspect_database 검색;

  • 설명 및 도구 JSON Schema에 8가지 작업 예시;

  • 일반 텍스트, JSON intent 또는 준비된 안전한 SELECT;

  • 모든 분석에서 cancelled 주문 제외;

  • 페이지네이션 및 결과 제한;

  • 다층 read-only 보호.

Related MCP server: Read-Only SQLite Shop Database MCP Server

설치

Python 3.11 이상이 필요합니다.

python3 -m venv .venv
.venv/bin/python -m pip install -e '.[dev]'

준비된 데이터베이스는 이미 data/shop.db에 있습니다. 150명의 고객이 있으며, customers.country 필드는 모든 고객에게 채워져 있고, 45명의 고객이 Germany에 속합니다.

데이터베이스 재준비

원본 데이터베이스는 변경되지 않습니다. 명령은 새 작업 복사본을 만들고 재현 가능한 합성 값으로 country TEXT NOT NULL을 추가합니다:

.venv/bin/shop-prepare-db /path/to/original/shop.db ./data/shop.db

분포는 customers.id에만 의존하므로 재준비 시 동일한 값이 생성됩니다. 국가는 테스트 데이터이며 고객의 실제 위치를 나타내지 않습니다.

경로 설정

경로는 다음 순서로 결정됩니다:

  1. SHOP_DB_PATH;

  2. SHOP_MCP_CONFIG 또는 config/shop_mcp.json의 JSON 파일;

  3. 작업 디렉터리 기준 data/shop.db.

환경 변수 예시는 .env.example에 있고, MCP 클라이언트 구성은 config/mcp.example.json에 있습니다. /absolute/path/to/MCPDeveloper를 프로젝트의 절대 경로로 바꾸세요.

실행

일반적으로 MCP 클라이언트가 자체 구성에서 서버를 실행합니다. 프로세스를 수동으로 실행하려면:

SHOP_DB_PATH=./data/shop.db .venv/bin/python -m shop_mcp.server

stdout은 MCP JSON-RPC용으로 예약되어 있습니다. 진단 정보는 프로토콜 스트림으로 출력되지 않습니다.

도구

inspect_database

인수:

{"action": "list_tables"}

또는:

{"action": "describe_table", "table": "customers"}

query_shop

인수:

{
  "request": "Сколько клиентов из Германии?",
  "limit": 100,
  "offset": 0
}

tools/list를 통해 선언되는 예시:

  1. 데이터베이스에 어떤 테이블이 있고 각 테이블에 어떤 필드가 있나요?

  2. 독일에서 온 고객은 몇 명인가요?

  3. 어느 국가에 고객이 가장 많나요?

  4. 어느 고객이 가장 많이 지출했나요? 이름, 이메일, 총 금액을 반환하세요.

  5. 판매 수량과 매출 기준 상위 5개 상품을 보여주세요.

  6. 매출 기준 상위 3개 카테고리를 보여주세요.

  7. 2025년 매출은 얼마였나요?

  8. 어느 고객이 가장 많은 주문을 했나요?

구조화된 intent를 전달할 수 있습니다:

{
  "request": "{\"intent\": \"revenue_by_year\", \"year\": 2025}"
}

또는 준비된 쿼리:

{
  "request": "SELECT name, category FROM products ORDER BY name"
}

비즈니스 규칙

  • orders.status = 'cancelled'는 지표에 포함되지 않습니다.

  • 고객 지출과 연간 매출은 orders.total_amount로 계산됩니다.

  • 상품 및 카테고리 매출은 quantity * unit_price로 계산됩니다.

  • 연도 간격은 반개방형입니다: 1월 1일 포함부터 다음 해 1월 1일 제외까지.

  • 제공된 데이터베이스의 모든 주문은 2026년에 속하므로 2025년 매출은 0입니다.

보안

서버는 mode=ro로 SQLite를 열고, query_only를 활성화하며, SQLite authorizer를 설정하고, 여러 문장, DML, DDL, ATTACH, 변경하는 PRAGMA, 확장 로드를 금지하며 실행 시간을 제한합니다.

SELECT가 아닌 작업을 실행하려는 모든 시도는 정확한 응답을 반환합니다:

Не доступный вариант запроса

SQLite 내부 오류, 스택 추적 및 로컬 경로는 클라이언트에 반환되지 않습니다.

검증

.venv/bin/pytest
.venv/bin/ruff check .
.venv/bin/ruff format --check .

테스트에는 MCP 서버를 subprocess로 실제 실행, stdio를 통한 ClientSession 연결, tools/list, 두 도구 모두 호출, 분석 시나리오 및 데이터베이스 변경 시도가 포함됩니다.

전체 승인된 사양은 docs/specification.md에 있습니다.

Available Tools

2 tools
inspect_databaseПросмотр структуры базы интернет-магазинаB

Возвращает список таблиц или поля выбранной таблицы. Используйте action=list_tables либо action=describe_table.

NameRequiredDescriptionDefault
tableNoНазвание таблицы для describe_table
actionYesДействие: list_tables или describe_table

Output Schema

NameRequiredDescription
resultYes

TDQS

B3.3/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries full responsibility for behavioral disclosure. It only states that it returns tables or fields, which is basic. It does not disclose any side effects, permission requirements, error behaviors, or what happens when the table does not exist. For a read-only inspection tool this is relatively benign, but the coverage is minimal.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single sentence that fully conveys the essential information. It is front-loaded with the primary function and immediately outlines the two actions. No redundancy or filler.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given that an output schema exists, the return format is likely covered elsewhere. The description is sufficient for a simple schema-inspection tool, but it lacks any mention of error handling or the distinction from query_shop. It is adequate but not comprehensive.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema already provides 100% coverage: 'action' is described as 'list_tables or describe_table' and 'table' as 'table name for describe_table'. The description repeats this mapping without adding new information, so it meets the baseline but does not exceed it.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's purpose: it returns a list of tables or fields of a selected table. This is a specific verb-resource combination and distinguishes it from the sibling query_shop, which presumably queries data rather than schema. The two modes (list_tables, describe_table) are explicitly mentioned.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides instructions on how to use the tool (choose action) but does not explain when to use this tool versus the sibling query_shop. It offers no context about typical scenarios, prerequisites, or exclusions. The usage guidance is limited to internal action selection, not selection among alternatives.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

query_shopЗапрос к базе интернет-магазинаA

Выполняет аналитический запрос к базе интернет-магазина только для чтения. MCP преобразует запрос в безопасный SQL SELECT. Отменённые заказы не учитываются. Примеры: 1) Какие таблицы есть в базе и какие в них поля? 2) Сколько клиентов из Германии? 3) В какой стране больше всего клиентов? 4) Какой клиент потратил больше всего? Верни имя, email и общую сумму. 5) Покажи топ-5 товаров по проданному количеству и выручке. 6) Покажи топ-3 категории по выручке. 7) Какая выручка была в 2025 году? 8) Какой клиент сделал больше всего заказов?

NameRequiredDescriptionDefault
limitNoМаксимальное число строк
offsetNoСмещение для пагинации
requestYesАналитический запрос обычным текстом, JSON intent или SQL SELECT.

Output Schema

NameRequiredDescription
resultYes

TDQS

A3.9/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries the full burden of behavioral disclosure. It explicitly states the tool is read-only, that MCP converts queries into safe SQL SELECT, and that canceled orders are excluded from results. This is useful and goes beyond minimal. It does not mention error handling or response format, but the core behavior is well disclosed.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is fairly long because it contains eight examples, but those examples are practical and illustrate valid request formats that are directly relevant to the tool. The purpose statement is front-loaded, and each example adds value, so the length is justified. It is well-structured and not overly verbose.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a natural-language query tool that has an output schema (indicated by the schema signals), the description covers essential context: what the tool does, how it handles queries, and what data is excluded (canceled orders). It could add more about limitations or error behavior, but given the output schema exists and the description is already thorough, it is largely complete.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100% — every parameter has a clear description. The description itself adds little beyond the schema: it repeats the request examples but does not elaborate on limit or offset semantics, which the schema already covers adequately. The baseline of 3 is appropriate because the schema handles the heavy lifting.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the verb and resource: 'performs a read-only analytical query to the online store database' and explains the MCP converts queries to safe SQL SELECT. It is specific about being read-only and analytical, which differentiates it from a generic database tool and gives a clear sense of its function, though it does not explicitly name the sibling tool.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Usage is implied through a set of concrete examples that show typical analytical questions (e.g., 'which country has most customers?'). However, the description does not explicitly state when to use this tool versus the sibling 'inspect_database', nor does it provide any exclusions or when-not-to-use guidance. It relies on the reader to infer that this is for read-only analytical queries.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

  1. 2 tool updatesv0.1.0
    • First observedinspect_database
    • First observedquery_shop
Deploy Server
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

A3.8/5.0

Scored across 2 tools

Disambiguation5/5

The two tools have clearly distinct purposes: one handles database schema inspection (list tables/describe fields), the other executes analytical queries. There is no overlap in their functional boundaries, making selection unambiguous for an agent.

Naming Consistency5/5

Both tools follow a consistent verb_noun snake_case pattern ('inspect_database', 'query_shop'), using clear, descriptive verbs. Naming style is uniform and predictable.

Tool Count3/5

With only 2 tools, the server feels minimal for a shop analytics domain. While the query tool is versatile, a more granular set (e.g., get_customers, get_orders) might be expected. The count is borderline, not excessive but thin.

Completeness5/5

The tool surface fully covers the stated purpose: schema discovery via inspect_database and arbitrary read-only analytical queries via query_shop. There are no obvious gaps for a read-only analytics server, as the query tool can address any data retrieval need.

Maintenance

ActivitySlowing
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers

  • Shop Analytics MCP Server

    bogdaamn
    F
    license
    A
    quality
    C
    maintenance
    Enables AI agents to answer analytical questions about an online store's SQLite database through specialized read-only tools, without any risk of modifying the underlying data.
    8
    -

  • Read-Only SQLite Shop Database

    alimbux
    F
    license
    A
    quality
    C
    maintenance
    Enables AI agents to safely inspect and query an SQLite e-commerce database with tools for listing tables, describing schemas, and running read-only SQL queries while blocking destructive operations.
    4
    -

  • shop-db

    aleksei-antipin
    F
    license
    A
    quality
    C
    maintenance
    Enables AI agents to read-only query an online store's SQLite database, listing tables, inspecting schemas, and running SELECT queries over customers, products, orders, and order items.
    3
    -

  • shop-mcp

    AndrewKonst
    F
    license
    Not graded
    quality
    C
    maintenance
    Enables AI agents to read-only analyze a SQLite e-commerce database, exploring schema and running analytical SQL queries over stdio.
    -