cdisc-mcp

cdisc-mcp

Exposes CDISC standards data including SDTM, ADaM, CDASH, and Controlled Terminology as tools for AI assistants via the CDISC Library API. It enables users to search standards, retrieve domain variables, and access codelist definitions to facilitate clinical research data management.

Category
访问服务器

README

cdisc-mcp

CDISC Library MCP Server — exposes CDISC standards data (SDTM, ADaM, CDASH, Controlled Terminology) as MCP tools that AI assistants can call directly.

For a full guide covering installation, API key setup, connecting to Claude Desktop or VS Code, tool reference, and troubleshooting, see the User Manual.

Installation

# Install with development dependencies (recommended)
pip install -e ".[dev]"

# Install runtime only
pip install -e .

Configuration

A CDISC Library personal API key is required. Obtain one at https://library.cdisc.org.

Linux / macOS

export CDISC_API_KEY=your_api_key_here

Windows (Command Prompt)

set CDISC_API_KEY=your_api_key_here

Windows (PowerShell)

$env:CDISC_API_KEY = "your_api_key_here"

Running the Server

# Via installed entry point
cdisc-mcp

# Via Python module
python -m cdisc_mcp.server

Available Tools (12)

Tool Description
list_products List all available CDISC standards and their published versions
search_cdisc Search across all CDISC standards by keyword
get_sdtm_domains List all SDTM domains for a given version
get_sdtm_domain_variables Get all variables defined in an SDTM domain
get_sdtm_variable Get the full definition of a specific SDTM variable
get_adam_datastructures List all ADaM data structures for a given version
get_adam_variable Get the definition of a specific ADaM variable
get_cdash_domains List all CDASH domains for a given version
get_cdash_domain_fields Get all data collection fields for a CDASH domain
list_ct_packages List all available Controlled Terminology packages
get_codelist Get a specific CT codelist definition and metadata
get_codelist_terms List all valid terms in a CT codelist (max 100)

Web Explorer (Optional)

A lightweight browser UI for interactively testing tools without an AI client.

# Install web dependencies
pip install -e ".[web]"

# Set API key and start the bridge server
set CDISC_API_KEY=your_api_key_here   # Windows CMD
python web/app.py

Open http://localhost:8080 in your browser. The explorer lets you select any of the 12 tools, fill in parameters, run them, and inspect syntax-highlighted JSON responses.

Development

Running Tests

# Run full test suite with coverage (must reach 80%)
pytest

# Run specific test files
pytest tests/test_config.py tests/test_client.py tests/test_response_formatter.py tests/test_tools.py -v

Project Structure

src/cdisc_mcp/
    __init__.py          # Package version
    client.py            # Async HTTP client with caching and retry
    config.py            # Configuration from environment variables
    errors.py            # Custom exception types
    response_formatter.py # Response cleaning and truncation
    server.py            # FastMCP server entry point
    tools/
        search.py        # list_products, search_cdisc
        sdtm.py          # SDTM domain and variable tools
        adam.py          # ADaM data structure and variable tools
        cdash.py         # CDASH domain and field tools
        terminology.py   # Controlled Terminology tools
        _validators.py   # Input validation helpers
web/
    app.py               # FastAPI bridge server (HTTP → tool functions)
    index.html           # Single-file browser explorer UI
tests/
    test_config.py
    test_client.py
    test_response_formatter.py
    test_tools.py
    test_errors.py
    test_server.py
    unit/                # Additional unit tests
    integration/         # Integration tests

Code Quality

# Linting
ruff check src/ tests/

# Type checking
mypy src/

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选