spear-mcp-test
MCP server for accessing SPEAR model output from various sources (AWS, STAC API, local) and integrating with AI assistants like Claude Desktop or a SPEAR Climate Chatbot.
README
Model Context Protocol (MCP) Test Server for Accessing SPEAR Model Output
Model Context Protocol is a standarized form of two-way communication for connecting AI applications to external services, devices and databases. This test MCP server is designed for accessing SPEAR output. There are 3 different 'flavors' of SPEAR MCP server labeled as different branches in this repository. Each branch is catered to a specified SPEAR output location which include AWS hosted, STAC API hosted and local mounted directory hosted files.
The SPEAR MCP server can be utilized in multiple ways. Currently, the most supported method is using the MCP server in parrallel with our SPEAR Climate Chatbot model. You can find this chatbot here:
https://github.com/zappalaja/spear-climate-chatbot
The MCP servers also work well with and have been tested using Claude Desktop and Claude Code.
This is not intended for operational purposes. Contact through GitHub issues for questions.
Running the MCP Server in Terminal
UV will need to be installed in the environment to run the MCP server.
You can download UV from here: https://github.com/astral-sh/uv
# On macOS and Linux. Run this in your chosen MCP env
curl -LsSf https://astral.sh/uv/install.sh | sh
To run the MCP server edit start_mcp_server.sh as needed, make it executable and run it in terminal:
chmod +x start_mcp_server.sh
./start_mcp_server.sh
Claude Desktop Install
To download Claude Desktop, follow the instructions from this link:
https://claude.ai/download
Once Claude Desktop is installed, go to the settings and proceed to the 'Developer' section. Click 'Edit config' and insert the correct form of the 'Local Install' into the 'claude_desktop_config.json'.
Configuration in LLM JSON file:
{
"mcpServers": {
"spear_mcp": {
"command": "uv", // May need a full path, like /opt/homebrew/bin/uv
"args": [
"--directory",
"/Absolute/path/to/spear-mcp",
"run",
"spear-mcp"
]
}
}
}
Claude Code Install
Claude Code Installation Page:
https://code.claude.com/docs/en/overview
One downloaded in terminal, run this command to connect to the SPEAR MCP server:
claude mcp add spear_mcp -- /bin/bash -c "cd /path/to/spear-mcp-test && uv run spear-mcp"
To check if you have successfully connected to the server, check with the "/mcp" command.
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。