MCP Map Server
A Model Context Protocol server that provides a dynamic map interface with tools to add raster/vector layers, control map view, and modify paint properties.
README
MCP Map Server
A Model Context Protocol (MCP) server that provides a dynamic map interface.

Quick Start
Add the hosted MCP endpoint to your LLM client, like so:
Using VSCode
create a .vscode/mcp.json like this: (as in this repo)
{
"servers": {
"map-server": {
"command": "uv",
"args": [
"run",
"mcp-map-server",
"--transport",
"stdio",
"--prompt-file",
"system-prompt.example.md"
]
},
}
}
You should see a little icon appear above the entry to "Start" server, like so:



Configuring Data Layer Information
The server supports MCP prompts to provide AI agents with information about available data layers. This helps agents understand which layers are available, their attributes, and how to use them effectively.
Recommended: Use a Markdown file
You can provide a system prompt as a plain markdown file:
mcp-map-server --prompt-file my-layers.md
Tooling & API
Every map-modifying tool returns the full, updated map configuration in JSON format.
Available Tools
add_layer: Adds a raster (XYZ/Tiles) or vector (MVT/PMTiles) layer.set_map_view: Moves the camera to a specificcenterandzoom.filter_layer: Applies MapLibre filter expressions to an existing layer.set_layer_paint: Dynamically modifies paint properties (colors, opacity, etc.).remove_layer: Deletes a layer by ID.get_map_config: Returns the current session/provided state as JSON.
MCP Client Configuration
To use this with MCP-compatible clients like Claude Desktop or VSCode extensions, point them to the server.
Local Development (Recommended):
Local clients like Claude Desktop and VSCode expect StdIO transport when launching a command. The server now defaults to stdio.
{
"mcpServers": {
"map-server": {
"command": "uv run mcp-map-server"
}
}
}
Passing CLI Arguments (Transport, Prompts, etc.):
If you need to specify the transport explicitly or provide a custom system prompt file, use the args array.
[!IMPORTANT] Arguments must be passed as individual elements in the
argsarray, NOT as part of thecommandstring.
{
"mcpServers": {
"map-server": {
"command": "uv run mcp-map-server",
"args": [
"--transport", "stdio",
"--prompt-file", "/path/to/my-layers.md"
]
}
}
}
Note: The server defaults to stdio when run as a command.
Using the Shared Production Server:
Web-based or remote clients connect via HTTP.
{
"mcpServers": {
"map-server": {
"url": "https://mcp-map.nrp-nautilus.io/mcp/"
}
}
}
License
MIT
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。