Mirobody
Lab results to LOINC codes and UCUM units, symptoms and diagnoses to ICPC-3. Offline, no API key.
作者提供的说明(英文原文)
Read-only Spring Boot runtime evidence for agents: beans, routes, health, startup.
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
你需要
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
打开 Claude Desktop → Settings → Developer → Edit Config,把下面的内容粘贴到 claude_desktop_config.json
{
"mcpServers": {
"spring-runtime": {
"command": "uvx",
"args": [
"spring-runtime-mcp"
]
}
}
}在终端中运行这条命令
claude mcp add spring-runtime -- uvx spring-runtime-mcp粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"spring-runtime": {
"command": "uvx",
"args": [
"spring-runtime-mcp"
]
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"spring-runtime": {
"type": "stdio",
"command": "uvx",
"args": [
"spring-runtime-mcp"
]
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
这个服务器在你的电脑上运行,而 ChatGPT 只能连接互联网上的服务器。请选择其他应用。
你的选择通过 Cookie 保存在这台设备上,无需账户。
这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
它在你的电脑上运行;安装后,工具列表会显示在你的 AI 应用中。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月6日
注册名称已通过 GitHub 账户 neilmorse 验证,该账户拥有此仓库
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
它在你的电脑上运行,没有互联网端点
在你的电脑上运行,改用本地设置
它在你的电脑上运行;我们只能读取互联网端点的工具列表
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在你的电脑上运行。你的 AI 应用会在需要时启动它,数据不会经过作者的服务器(除非它自己调用在线服务)。
GitHub 仓库最后更新于–,最新版本(0.2.12)于4天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。你的电脑上还需要有 Python (uv)。
同一分类中有:Mirobody, FHIR MCP Server, Google Health MCP
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
Lab results to LOINC codes and UCUM units, symptoms and diagnoses to ICPC-3. Offline, no API key.
MCP server providing seamless access to FHIR APIs for AI tools and healthcare applications
Privacy-first unofficial Google Health API v4 MCP server for health, sleep, activity, HR agents
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
Chinese README
Give your coding assistant access to what is actually running in a Spring Boot application. This MCP server reads authorized Actuator and OpenAPI data to help investigate Beans, routes, health, database migrations and startup problems.
| Your question | What it provides |
|---|---|
| Why was a Bean not created? | Beans, dependencies and condition reports, with blocking chains when evidence is sufficient |
| Is this route registered? | Actual paths, HTTP methods, Controllers and MVC/WebFlux mappings |
| What is unhealthy? | Visible components and liveness/readiness status |
| Did a migration fail? | Reported Flyway/Liquibase records, states and checksums |
| What made startup slow? | Slow steps and their parent chains |
| Where did a configuration value come from? | Sources for registered properties; configuration diagnostics require opt-in |
| Is this API in OpenAPI? | Operations, schemas and references within the document |
| Does the process match my expectations? | Compare the Bean, route, migration version and profile expectations you supply |
For example, ask: “Check whether GET /orders/{id} is registered and tell me which Controller handles it.” The server reads mapping metadata without calling the business endpoint.
You need Python 3.11+, uv and an authorized local Spring Boot application. Confirm its management endpoints and permissions using the target setup guide, then check and start 0.2.12 Alpha:
uvx --default-index https://pypi.org/simple spring-runtime-mcp==0.2.12 --version
uvx --default-index https://pypi.org/simple spring-runtime-mcp==0.2.12 --base-url http://127.0.0.1:18081
Use the second command as your MCP client's STDIO server command. In a terminal it waits for client requests. Follow the English first-query guide; additional client configuration templates cover other clients.
Try the synthetic route demo using the public package without preparing an application, then share first-query feedback. The replay needs no Spring process and does not establish live application or client compatibility.
The server connects only to loopback addresses, uses fixed endpoints and supports optional Bearer authentication from a named environment variable. OpenAPI uses /actuator/openapi. Configuration diagnostics are disabled by default and require explicit opt-in via CLI flags (--config-prefix, --config-property, --config-profiles). When enabled, values remain hidden unless both --config-allow-values and the tool’s includeValues=true are set. Output is sanitized, and each tool response is capped at 16 KiB. Missing evidence is reported as partial/unknown.
Tools do not write to the target, restart applications, or read heap dumps or raw logs. The complete Boot/Java/client matrix, Windows/macOS and production-cluster compatibility remain unverified.
Documentation · Tool contract · Source facts · Development and checks
Apache-2.0 · Contributing · Report a security issue