mcp-server-redcap

mcp-server-redcap

Enables interaction with REDCap instances for managing research data, projects, and files through natural language.

Category
访问服务器

README

mcp-server-redcap

A Model Context Protocol (MCP) server for interacting with REDCap (Research Electronic Data Capture) instances.

Exposes 38 tools covering data access, project design, file management, and longitudinal project structure. Designed for use with LLM agents that need to query or build REDCap projects in natural language conversations.

Tools

Records

Tool Description
export_records Export records with optional filters by field, record ID, or event
import_records Import records from a JSON payload
delete_records Delete records by ID
export_report Export a saved REDCap report by its ID
export_field_names Export the mapping from field names to export names (required for checkbox fields)
generate_next_record_name Get the next available record ID respecting the project's auto-numbering

Metadata

Tool Description
get_project_info Retrieve project-level information and settings
get_metadata Export the data dictionary, optionally filtered by field or form
export_logging Export the project audit log, with filters by type, user, record, or date range
export_repeating_instruments_events Export the repeating instruments/events configuration
import_repeating_instruments_events Update the repeating instruments/events configuration

Instruments & project design

Tool Description
get_instruments List all instruments with internal names and display labels
get_instrument_event_mappings List arm/event/instrument mappings (longitudinal projects)
add_field Add a new field to an instrument (round-trip through data dictionary)
remove_field Remove a field from the data dictionary
update_field Update specific properties of a field without touching the rest
add_instrument Create a new instrument
delete_instrument Delete an instrument and all its fields
rename_instrument Rename an instrument's internal name
move_field Reposition a field within the data dictionary
clone_instrument Copy all fields of an instrument into a new one with a field-name prefix
assign_instrument_to_event Add an instrument to a longitudinal event
unassign_instrument_from_event Remove an instrument from a longitudinal event

Analysis

Tool Description
get_project_structure Compact overview of the whole project: instruments, field counts, event matrix
check_field_references Find all fields that reference a given field in branching logic or calculations
validate_branching_logic Scan all branching logic for references to non-existent fields

Files

Tool Description
export_file Download a file attachment from a record field (returned as base64)
import_file Upload a file to a record field (accepts base64 content)
delete_file Delete a file attachment from a record field
export_pdf Export a PDF of one or more instruments for a record (returned as base64)

Arms & events (longitudinal)

Tool Description
export_arms List arms
import_arms Create or update arms
delete_arms Delete arms by number
export_events List events, optionally filtered by arm
import_events Create or update events
delete_events Delete events by unique name

Surveys

Tool Description
export_survey_link Get a participant-specific survey URL for a record
export_survey_participant_list List survey participants and their response status

Installation

pip install mcp-server-redcap

Or with uv:

uvx mcp-server-redcap

REDCap version compatibility

The server works against any REDCap instance running version 8.0 or later. Most tools are available from version 6.x, but export_pdf requires 8.x. Repeating instruments (export_repeating_instruments_events, import_repeating_instruments_events) require 6.16+. Tools that are unavailable on a given instance return a plain error string rather than crashing.

The current REDCap version of the connected instance is included in the output of get_project_structure.

Configuration

The server reads connection details from environment variables:

Variable Required Description
REDCAP_URL Yes Full REDCap API endpoint URL (e.g. https://redcap.example.org/api/)
REDCAP_TOKEN Yes Project-level API token from REDCap → API → Generate Token
REDCAP_VERIFY_SSL No Set to false to skip SSL verification (default: true)

Create a .env file in your working directory and the server will load it automatically:

REDCAP_URL=https://redcap.example.org/api/
REDCAP_TOKEN=your_project_token_here

Usage with Claude Code

Add the server to your Claude Code MCP configuration:

claude mcp add redcap -e REDCAP_URL=https://redcap.example.org/api/ \
                      -e REDCAP_TOKEN=your_token -- uvx mcp-server-redcap

Or edit .claude/settings.json manually:

{
  "mcpServers": {
    "redcap": {
      "command": "uvx",
      "args": ["mcp-server-redcap"],
      "env": {
        "REDCAP_URL": "https://redcap.example.org/api/",
        "REDCAP_TOKEN": "your_project_token_here"
      }
    }
  }
}

Development

git clone https://github.com/msicilia/mcp-server-redcap
cd mcp-server-redcap
uv sync --extra dev

Run the MCP inspector for interactive local testing:

uv run mcp dev src/mcp_server_redcap/server.py

Run the server directly (stdio transport, for use with MCP clients):

uv run mcp-server-redcap

Project structure

src/mcp_server_redcap/
├── __init__.py
├── __main__.py          # python -m mcp_server_redcap
├── server.py            # FastMCP server factory and entry point
├── connection.py        # REDCap project connection (lazy, singleton)
└── tools/
    ├── __init__.py
    ├── records.py       # record CRUD, field names, next record name
    ├── metadata.py      # project info, data dictionary, logging, repeating instruments
    ├── instruments.py   # instrument listing, field and instrument mutations
    ├── analysis.py      # project structure overview, reference and logic validation
    ├── files.py         # file attachments and PDF export
    ├── arms_events.py   # longitudinal arms and events CRUD
    └── surveys.py       # survey links and participant lists

See DESIGN.md for the full API reference and design rationale.

License

MIT

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