picgo-uploader
Enables uploading images to a running PicGo application through its built-in server.
README
PicGo Uploader MCP Server
An MCP (Model Context Protocol) server that allows interaction with a running PicGo application to upload images.
Overview
This server exposes PicGo's image uploading capabilities as an MCP tool. It connects to the PicGo application's built-in server (usually running on http://127.0.0.1:36677) and provides a tool to upload local image files.
Prerequisites
- Node.js: Ensure you have Node.js (v18 or later recommended) installed.
- PicGo Application: You need the PicGo desktop application installed and running.
- PicGo Server Enabled: In PicGo's settings (
PicGo 设置->设置Server), enable the server. The default port is36677. Make sure it's running. - NPM or Yarn: For installing the package.
Installation
npm install -g picgo-uploader
or
yarn global add picgo-uploader
This will install the MCP server globally and make the picgo-uploader command available.
Running the Server
Simply execute the command in your terminal:
picgo-uploader
The server will start and listen for MCP connections via stdio. You should see a message like:
PicGo Uploader MCP server running on stdio
Keep this terminal window open while you want to use the server.
Usage (MCP Tool)
Once the server is running and connected to your MCP client (like Roo Code), you can use the provided tool:
Tool: upload_image_via_picgo
- Description: Uploads one or more images using the running PicGo server application.
- Input Schema:
{ "type": "object", "properties": { "image_paths": { "type": "array", "items": { "type": "string" }, "description": "An array of absolute paths to the image files to upload." } }, "required": ["image_paths"] } - Output: On success, returns a JSON string containing the array of uploaded image URLs provided by PicGo. On failure, returns an error message.
Example (use_mcp_tool):
<use_mcp_tool>
<server_name>picgo-uploader</server_name>
<tool_name>upload_image_via_picgo</tool_name>
<arguments>
{
"image_paths": [
"C:\\Users\\YourUser\\Pictures\\screenshot1.png",
"/home/user/images/diagram.jpg"
]
}
</arguments>
</use_mcp_tool>
Note: Ensure the file paths provided in image_paths are absolute paths accessible from the machine where the picgo-uploader server is running.
Troubleshooting
- Error: "PicGo server request error: connect ECONNREFUSED 127.0.0.1:36677":
- Make sure the PicGo application is running.
- Verify that the PicGo server is enabled in its settings and running on the default port
36677.
- Error: "Image path does not exist: <path>":
- Double-check that the provided image path is correct and absolute.
- Ensure the file exists at the specified location on the machine running the
picgo-uploader.
- Error: "PicGo upload failed: ...":
- This indicates an error reported by PicGo itself (e.g., invalid configuration for the selected uploader, network issues). Check the PicGo application logs for more details.
Development
- Clone the repository.
- Install dependencies:
npm install - Build the code:
npm run build - Run in development:
node build/index.js - Watch for changes:
npm run watch(in a separate terminal)
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。
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。
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。
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。
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。
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。
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。
mcp-server-qdrant
这个仓库展示了如何为向量搜索引擎 Qdrant 创建一个 MCP (Managed Control Plane) 服务器的示例。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。