What is Vscode Mcp Feedback Extension

mcp-feedback is a Visual Studio Code extension designed to facilitate user feedback collection by integrating with an external AI assistant. It allows users to submit their feedback and receive responses via a dedicated interface within the VSCode environment.

Use cases

This extension can be used by developers and teams to gather user input on code features, bug reports, or general suggestions. It enhances the interaction between users and the development process, enabling more collaborative and informed software improvements.

How to use

To use the mcp-feedback extension, users can either install it directly from the VSCode marketplace by searching for ‘mcp-feedback’ or manually download the .vsix file and install it through the command palette. After installation, the extension is activated, allowing users to input feedback in the sidebar, which the AI assistant will process.

Key features

Key features of mcp-feedback include an intuitive user interface for submitting feedback, real-time interaction with an AI assistant, and customizable server configurations via a cursor configuration file. The integration allows for efficient feedback management directly within the coding environment.

Where to use

mcp-feedback is primarily used within Visual Studio Code, a popular code editor among developers. It is particularly useful in collaborative environments where continuous feedback is essential for iterative development and project management.

mcp-feedback (VSCode Extension)

功能

安装

1. 在VSCode中按 Ctrl+Shift+P 或 Cmd+Shift+P
2. 输入 “Extensions: Install from VSIX…”
3. 选择 .vsix 文件安装

使用

商店搜索安装

1. 在VSCode中按 Ctrl+Shift+P 或 Cmd+Shift+P
2. 输入 “Extensions: Search in Marketplace…”
3. 搜索 “mcp-feedback” 并安装

手动安装

1. 下载 .vsix 文件
2. 在VSCode中按 Ctrl+Shift+P 或 Cmd+Shift+P
3. 输入 “Extensions: Install from VSIX…”
4. 选择 .vsix 文件安装

激活插件

1. 插件激活后，左侧出现 “mcp-feedback” 图标
2. 外部AI助手通过 http://localhost:7423/mcp 调用MCP工具
3. 用户在侧边栏输入反馈，结果返回给AI助手

cursor配置文件

在 .cursor/mcp.json 添加：

{
  "mcpServers": {
    "mcp-feedback": {
      "url": "http://localhost:7423/mcp"
    }
  }
}