Vista API MCP Server
Provides secure access to VA's VistA electronic health record system, enabling AI applications to retrieve and summarize patient medical records for clinical staff.
README
Welcome to Vista API MCP Server
This repository contains documentation about a Model Context Protocol (MCP) server that provides secure access to VA's VistA electronic health record (EHR) system, enabling AI applications to retrieve and summarize patient medical records for clinical staff.
Overview
The Vista API MCP Server enables Large Language Models (LLMs) to securely access VistA by acting as a bridge that safely connects LLMs to patient medical records and clinical documentation. The server can be consumed by multiple AI applications, like AI Assist, to perform summarization or other LLM-enabled tasks.
What is an MCP server?
Model Context Protocol (MCP) servers follow a standardized protocol to expose resources, tools, and prompts to AI applications. They act as bridges that allow AI applications to access external capabilities without each application needing to implement custom integrations.
In this case, our MCP server converts VistA's legacy data formats for contemporary AI applications. VistA contains large amounts of patient data that is difficult to access programmatically. The MCP server allows authorized AI applications to access this rich clinical data without requiring each application to understand VistA's specific database structure or APIs.
Healthcare data is notoriously complex—lab results, medication histories, diagnostic imaging, and clinical notes all have different formats. The MCP server standardizes how this data is presented to AI applications, making it easier to build clinical tools.
How is an MCP server different from an LLM?
The MCP server acts as a secure data gateway that retrieves specific patient information from VistA and formats it for an LLM. The LLM can then process and generate human-like text responses based on prompts. The distinction is that the LLM does not directly access VistA or understand medical record formats, the MCP server is the bridge.
What’s the value of this MCP server?
Without an MCP server, teams would need to build VistA integration directly into each AI application, leading to technical debt, security risks, and development inefficiencies.
Whereas, any application that supports Model Context Protocol can consume this server. Think of it as a reusable data service that multiple products can leverage while maintaining strict security boundaries. For example, a medication management tool and AI Assist could both use the same MCP server to access patient data without each building their own VistA integration. This approach is faster, more secure, and creates reusable infrastructure that accelerates future VA AI initiatives.
Roadmap
Our roadmap provides an overview of how we plan to build this MCP server.

Review our roadmap in VA Mural.
Get started
- Development guide - Setup instructions and development workflows
- Deployment guide - Production deployment instructions
- Test data guide - Test patient IDs and sample data
- Logging guide - Logging configuration and best practices
- Troubleshooting guide - Common issues and solutions
- Data models reference - Complete data structure documentation
Connect with us
Reach out on Office of CTO Slack: #va-ai-chat-public
推荐服务器
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。
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。