How do I use kfda-mcp?

1. Click on "Install Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@kfda-mcp check DUR interaction between ibuprofen and aspirin" That's it! The server will respond to your query, and you can continue using it as needed. Here is a step-by-step guide with screenshots.

kfda-mcp

by pianovirus

Overview Schema Related Servers Score Discussions

Python

Remote

🇰🇷 KFDA MCP Server

MCP (Model Context Protocol) server for Korean Ministry of Food and Drug Safety (식약처) public OpenAPI.

LLM agents (Claude, ChatGPT, etc.) can autonomously query Korean drug master, DUR (Drug Utilization Review) safety rules, and health supplement databases through the standard MCP protocol.

DOI

🎯 Why this exists

The Korean healthcare AI ecosystem lacks an open MCP server for MFDS (식약처) public APIs. Global MCP servers like BioMCP focus on US data (FDA, PubMed). For Korean healthcare AI applications — pharmacy systems, telemedicine, drug interaction checkers — direct access to KFDA/MFDS APIs is essential but each team has to write their own wrapper.

This project provides a standard MCP server so any LLM agent can query Korean drug safety data without bespoke integration code.

Related MCP server: pykrx-mcp

✨ Features

Tool	Description	KFDA OpenAPI
`search_drug`	의약품 마스터 검색 (제품명·성분·제조사·ATC 코드)	의약품 제품 허가정보
`check_dur_interaction`	두 약물의 병용금기·연령금기·임부금기 확인	의약품 안전사용서비스 (DUR)
`search_supplement`	건강기능식품 인허가 정보 검색	식품안전나라 건기식 OpenAPI
`get_drug_easy_info`	e약은요 (환자용 쉬운 약물 정보) 조회	e약은요
`lookup_atc_code`	ATC 코드 → 약물군 분류 정보	내장 매핑

🚀 Quick Start

1. Install

pip install kfda-mcp
# or from source
git clone https://github.com/pianovirus/kfda-mcp
cd kfda-mcp
pip install -e .

2. Get your MFDS API key

Apply for a free API key at data.go.kr (공공데이터포털) — select 식약처 OpenAPI services.

3. Configure

Create .env:

MFDS_API_KEY=your_api_key_here

4. Run as MCP server

kfda-mcp

5. Connect from Claude Desktop

Add to ~/.config/claude/claude_desktop_config.json (or %APPDATA%\Claude\claude_desktop_config.json on Windows):

{
  "mcpServers": {
    "kfda": {
      "command": "kfda-mcp",
      "env": {
        "MFDS_API_KEY": "your_api_key_here"
      }
    }
  }
}

Restart Claude Desktop. Now Claude can call Korean drug safety tools autonomously.

💡 Example Usage

Ask Claude (or any MCP-compatible LLM):

"환자가 와파린을 복용 중인데 자몽 추출 영양제를 같이 먹어도 되는지 확인해줘"

Claude will autonomously:

Call search_drug("와파린") → get warfarin info + ATC code
Call check_dur_interaction("warfarin", "grapefruit_extract") → DUR check
Synthesize a safety recommendation grounded in MFDS data

🏗 Architecture

LLM Agent (Claude / GPT / etc.)
    │
    │ MCP Protocol (stdio / SSE)
    │
    ▼
KFDA MCP Server (this project)
    │
    │ HTTP requests
    │
    ▼
MFDS Public APIs
  - 의약품 제품 허가정보 OpenAPI
  - 의약품 안전사용서비스 (DUR)
  - 식품안전나라 건기식 OpenAPI
  - e약은요

📋 Tool Reference

`search_drug(name: str, limit: int = 10)`

Search Korean drug master by product name, generic name, or manufacturer.

Returns: List of drugs with product_name, generic_name, manufacturer, atc_code, kor_indication, dosage_form.

`check_dur_interaction(drug_a: str, drug_b: str)`

Check DUR safety rules between two drugs (병용금기·연령금기·임부금기 등).

Returns: List of warnings with rule_type, severity, description_kor.

`search_supplement(ingredient: str | None = None, product: str | None = None)`

Search 식약처 건강기능식품 (functional health food) registrations.

Returns: List of supplements with product_name, manufacturer, main_ingredient, approved_function.

`get_drug_easy_info(drug_name: str)`

Get patient-friendly drug information from e약은요 (consumer-facing drug guide).

Returns: purpose, dosage, side_effects, warnings in Korean.

`lookup_atc_code(atc_code: str)`

Look up ATC (Anatomical Therapeutic Chemical) classification by code.

Returns: level1 ~ level5 Korean & English names, drug examples.

🛣 Roadmap

Project scaffold + MCP SDK setup
search_drug — 의약품 마스터 ✅ live verified (타이레놀 → ATC N02BE01)
check_dur_interaction — DUR 병용금기 ✅ live verified (아스피린+와파린 → 50건)
search_supplement — 건기식 ✅ live verified (식품안전나라 C003, 비타민 검색 OK)
get_drug_easy_info — e약은요 ✅ live verified (타이레놀 효능·용법·부작용 반환)
lookup_atc_code — ATC 매핑 (planned)
Local cache layer (reduce API calls) — planned
Async batch query support — planned
Multilingual responses (한/영) — planned

현재 상태: 4/4 핵심 tool 모두 live 검증 완료, production-ready

Local cache layer (reduce API calls)
Async batch query support
Multilingual responses (한/영)

🤝 Contributing

Korean healthcare AI engineers — contributions welcome! Please open an issue or PR.

This project aims to be the de facto MCP server for MFDS public APIs.

📜 License

🙏 Acknowledgments

Anthropic MCP — Standard protocol for LLM-tool communication
식품의약품안전처 (MFDS) — Public health data API provider
공공데이터포털 — Korean government open data platform

한국어 요약

식약처(MFDS)의 공공 OpenAPI를 MCP 프로토콜로 감싼 서버입니다. Claude, GPT 같은 LLM 에이전트가 한국 의약품 마스터·DUR 안전성·건강기능식품 정보를 자율적으로 조회할 수 있습니다.

한국 헬스 AI 개발자들이 매번 식약처 API wrapper를 직접 만들 필요 없이, 이 MCP 서버 하나로 표준화된 도구 접근을 제공하는 것이 목표입니다.

기여 환영합니다. 🌱

Available Tools

4 tools

check_dur_interactionB

식약처 의약품 안전사용서비스(DUR)에서 두 약물의 병용금기·연령금기 등 안전성 정보를 확인합니다. Check Korean Drug Utilization Review (DUR) safety rules between two drugs.

ParametersJSON Schema

Name	Required	Description	Default
`drug_a`	Yes	약물 A 이름 또는 성분명
`drug_b`	Yes	약물 B 이름 또는 성분명

TDQS

B3.2/5.0

Behavior2/5

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

No annotations provided, and description only states the tool checks safety rules. Does not disclose behavior for invalid inputs, error handling, or any limitations.

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?

Description is concise with two short sentences, but the Korean and English versions are redundant. Front-loads the purpose.

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?

For a simple query tool with full schema coverage, the description is adequate but does not mention return format or response behavior.

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 coverage is 100% with descriptions for both parameters. The tool description adds no additional semantic value beyond the schema.

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 checks Korean DUR safety rules between two drugs, specifying contraindications and age restrictions. It is distinct from sibling tools which focus on drug info and search.

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?

No guidance on when to use this tool vs siblings. Does not mention scenarios or prerequisites for use.

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

get_drug_easy_infoA

식약처 e약은요 (환자용 쉬운 약물 정보)를 조회합니다. Get patient-friendly Korean drug information: purpose, dosage, side effects, warnings.

ParametersJSON Schema

Name	Required	Description	Default
`drug_name`	Yes	약품명

TDQS

A3.6/5.0

Behavior2/5

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

No annotations are provided, so the description must carry the behavioral burden. It states the tool retrieves information (read-only implied) but does not disclose any behavioral traits such as authentication needs, rate limits, data source scope, or what happens on missing or invalid input. The description is insufficient for a mutation-free query tool.

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?

Two concise sentences (one Korean, one English) that efficiently convey the tool's purpose and content. No wasted words or redundancy. The description is front-loaded and easy to parse.

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?

With one required parameter and no output schema, the description covers the essential return content (purpose, dosage, side effects, warnings). It does not specify the response format (e.g., JSON structure), error handling, or edge cases, but for a simple lookup tool, the information is largely adequate.

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 coverage is 100% with one parameter (drug_name) described as '약품명'. The description adds that the drug name should be in Korean and that the information is patient-friendly, but does not provide additional semantics like expected format, supported languages, or examples. Baseline 3 is appropriate since the schema already covers the parameter.

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 explicitly states it queries patient-friendly Korean drug information and lists specific content categories (purpose, dosage, side effects, warnings). It clearly distinguishes from sibling tools like check_dur_interaction (drug interactions) and search_drug (general search) by focusing on easy-to-understand patient info.

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?

The description implies the tool should be used when seeking simplified patient drug information, but it does not explicitly state when to use this tool versus alternatives like search_drug or check_dur_interaction. Sibling names provide some context but no direct guidance.

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

search_drugA

식약처 의약품 마스터에서 약품을 검색합니다 (Korean drug master search). 제품명, 성분명, 제조사 등으로 조회 가능. Returns product_name, generic_name, manufacturer, ATC code, indication.

ParametersJSON Schema

Name	Required	Description	Default
`name`	Yes	검색할 약품명 (한글 또는 영문, 부분 매치 지원)
`limit`	No	반환할 최대 결과 수

TDQS

A4/5.0

Behavior3/5

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

No annotations are provided, so the description carries the full burden. It discloses return fields and mentions partial match support, but lacks details on auth needs, rate limits, or behavior for no results. Adequate but not rich.

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 concise: two sentences conveying purpose, searchable fields, and return fields. Front-loaded with the main action and no extraneous information.

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 simple search tool with 2 parameters, the description and schema together cover the essential aspects (parameters, return fields). No output schema but return fields are listed. Lacks details on error handling or pagination, but acceptable for this complexity.

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

Parameters4/5

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

Schema covers both parameters with descriptions (100% coverage). The description adds value by listing what can be searched (product name, ingredient, manufacturer), extending the understanding of the 'name' parameter beyond the schema's description.

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 searches a Korean drug master and lists searchable fields (product name, ingredient, manufacturer) and return fields. It distinguishes from siblings like search_supplement and check_dur_interaction by specifying the domain.

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?

The description implies usage for drug searches but does not provide guidance on when not to use it or compare to alternatives. Sibling tools are listed but no explicit when-to-use or when-not-to-use context.

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

search_supplementA

식약처 식품안전나라 건강기능식품 인허가 정보를 검색합니다. Search Korean functional health food (건기식) registrations by ingredient or product name.

ParametersJSON Schema

Name	Required	Description
`limit`	No	반환할 최대 결과 수
`product`	No	제품명
`ingredient`	No	주성분명 (예: 코엔자임Q10, 비타민D)

TDQS

A4.1/5.0

Behavior3/5

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

With no annotations provided, the description carries the full burden. It discloses the tool does a search (read operation) and specifies search fields, but does not mention pagination, result format, or any side effects. This is adequate for a simple lookup but lacks depth.

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 two sentences (Korean and English), directly stating purpose and search criteria. No unnecessary words, and the information is front-loaded.

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 the tool has no output schema and no required parameters, the description should at least indicate what is returned (e.g., list of registrations, details). It does not, leaving the agent uncertain about the output format. However, for a simple search, it is moderately complete.

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

Parameters4/5

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

Schema description coverage is 100%, so the schema already documents parameters. The description adds value by providing an example ingredient list ('코엔자임Q10, 비타민D'), which helps the agent understand valid values beyond the schema description.

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 searches Korean functional health food registrations by ingredient or product name. It uses a specific verb ('search') and resource ('건강기능식품 인허가 정보'), and distinguishes itself from sibling tools which focus on drug-related information.

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

Usage Guidelines4/5

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

The description implies usage when needing to find health food registrations, and specifies search criteria (ingredient or product name). While no explicit when-not or alternatives are given, the sibling tools are clearly different domains, so the usage context is clear.

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

Install Server

RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

TDQS

A3.9/5.0

Disambiguation5/5

Each tool has a clearly distinct purpose: interaction checking, patient info retrieval, drug search, and supplement search. There is no overlap or ambiguity among them.

Naming Consistency5/5

All tool names follow a consistent verb_noun pattern with underscores (e.g., check_dur_interaction, search_drug). The naming is predictable and uniform.

Tool Count5/5

With 4 tools, the server is well-scoped for a focused domain of Korean drug and supplement information. Each tool serves a distinct function without redundancy.

Completeness4/5

The server covers drug search, patient information, interaction checking, and supplement search—key functionalities. Minor gaps exist (e.g., professional-level drug details), but the core workflows are supported.

Maintenance

ActivityStale

ResponsivenessSyncing

Resources

GitHub Repository

Need Help?

Related Servers

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

mcpweave
com.mcpweave
Korea-native MCP gateway: Korean commerce, payments, messaging, gov & finance APIs for AI agents.
E-commerce & Retail Payments & Billing Finance
2
Airbyte Agents
io.github.airbytehq
Let AI agents query data and act across all your business apps via MCP.
Data Platforms AI & Machine Learning RAG Systems
2
Cralo
ai.cralo
Pay-per-use tool marketplace for AI agents. Search, price-check, and call APIs via MCP.
Agent Orchestration Developer Tools
1
kr-groundtruth A
io.github.ChloePark85
Korean fact-verification tools for AI agents: business registration, address, DART, apt prices, laws
Government Data Open Data Search
81

Related MCP Servers

KPIC MCP Server
Health & Wellness Search Research & Data
antegral
A
license
A
quality
A
maintenance
Enables users to search and retrieve detailed pharmaceutical drug information from the Korea Pharmaceutical Information Center (KPIC) API. Supports drug name searches and detailed lookups by drug code including ingredients, usage, precautions, and storage information.
Updated a month ago (2026-07-27 23:24 UTC)
2
5
MIT
pykrx-mcpofficial
Finance Research & Data
sharebook-kr
A
license
Not graded
quality
C
maintenance
Provides Korean stock market data (KOSPI, KOSDAQ, KONEX) including prices, fundamentals, investor trading, short selling, and indices via MCP protocol, enabling natural language queries from AI agents like ChatGPT and Claude.
Updated 7 months ago (2026-02-01 02:52 UTC)
3
MIT
korea-health-mcp-server
Government Data Open Data Biology & Medicine
NSLC7779
A
license
A
quality
C
maintenance
Enables natural language querying of Korean public medical data (data.go.kr) through MCP clients like Claude, providing tools for searching hospitals, drugs, non-payment costs, disease statistics, and more.
Updated 2 months ago (2026-06-23 01:15 UTC)
10
1
MIT
pubdata-mcp
Government Data Open Data
jcyeom
F
license
A
quality
C
maintenance
Enables querying Korean public datasets (apartment prices, weather, air quality) via natural language using an MCP server and local LLM agent.
Updated a month ago (2026-07-31 08:09 UTC)
5

Latest Blog Posts

Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)
By Om-Shree-0709 on July 25, 2026.
mcp
Agent Identity
OAuth 2.1
Your AI Chatbot Just Exposed Your CEO's Salary to an Intern
By Om-Shree-0709 on July 2, 2026.
Agent Identity
MCP Security
OAuth Delegation
Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)
By Om-Shree-0709 on June 30, 2026.
Agentic Ai
Prompt Injection
WebAssembly

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/pianovirus/kfda-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server

🇰🇷 KFDA MCP Server

🎯 Why this exists

✨ Features

🚀 Quick Start

1. Install

2. Get your MFDS API key

3. Configure

4. Run as MCP server

5. Connect from Claude Desktop

💡 Example Usage

🏗 Architecture

📋 Tool Reference

search_drug(name: str, limit: int = 10)

check_dur_interaction(drug_a: str, drug_b: str)

search_supplement(ingredient: str | None = None, product: str | None = None)

get_drug_easy_info(drug_name: str)

lookup_atc_code(atc_code: str)

🛣 Roadmap

🤝 Contributing

📜 License

🙏 Acknowledgments

한국어 요약

Available Tools

Maintenance

Resources

Looking for Admin?

Related MCP Connectors

Related MCP Servers

pykrx-mcpofficial

Latest Blog Posts

MCP directory API

`search_drug(name: str, limit: int = 10)`

`check_dur_interaction(drug_a: str, drug_b: str)`

`search_supplement(ingredient: str | None = None, product: str | None = None)`

`get_drug_easy_info(drug_name: str)`

`lookup_atc_code(atc_code: str)`