Skip to main content
Glama
Sign In
mcp-feedback-ultra

mcp-feedback-ultra

by YangHang0210
OverviewSchemaSourceRelated ServersDiscussionsAdmin
JavaScript

MCP Feedback Ultra

MCP Feedback Ultra 是一个用于 AI 辅助开发的交互式反馈服务器，支持 Web UI 和桌面应用双界面模式。

原始项目: noopstudios/interactive-feedback-mcp

核心特性

  • 双界面支持：Web UI（浏览器）+ 桌面应用（原生应用）

  • 智能环境检测：自动识别 SSH Remote、WSL 等特殊环境

  • 交互式反馈：AI 完成任务后主动收集用户反馈

  • 跨平台支持：Windows、macOS、Linux

Related MCP server: mcp-feedback-enhanced-gw

快速开始

安装

uvx mcp-feedback-ultra@latest

MCP 配置

在 Cursor 的 MCP 配置文件中添加：

{
  "mcpServers": {
    "mcp-feedback-ultra": {
      "command": "uvx",
      "args": ["mcp-feedback-ultra@latest"],
      "timeout": 600,
      "autoApprove": ["interactive_feedback"]
    }
  }
}

环境变量（可选）

变量

说明

默认值

MCP_DEBUG

调试模式

false

MCP_WEB_HOST

Web UI 绑定地址

127.0.0.1

MCP_WEB_PORT

Web UI 端口

8765

MCP_DESKTOP_MODE

启用桌面应用模式

false

使用方式

Web UI 模式（默认）

{
  "mcpServers": {
    "mcp-feedback-ultra": {
      "command": "uvx",
      "args": ["mcp-feedback-ultra@latest"]
    }
  }
}

桌面应用模式

{
  "mcpServers": {
    "mcp-feedback-ultra": {
      "command": "uvx",
      "args": ["mcp-feedback-ultra@latest"],
      "env": {
        "MCP_DESKTOP_MODE": "true"
      }
    }
  }
}

开发

本地测试

# 克隆项目
git clone https://github.com/YangHang0210/mcp-feedback-ultra.git
cd mcp-feedback-ultra

# 安装依赖
uv sync --dev

# 测试 Web UI
uv run python -m mcp_feedback_ultra test --web

# 测试桌面应用
uv run python -m mcp_feedback_ultra test --desktop

常见问题

SSH Remote 环境无法访问？

设置 MCP_WEB_HOST 为 0.0.0.0 允许远程访问：

{
  "env": {
    "MCP_WEB_HOST": "0.0.0.0",
    "MCP_WEB_PORT": "8765"
  }
}

然后在本地浏览器访问：http://[远程IP]:8765

许可证

MIT License

链接

Available Tools

2 tools
get_system_infoA

獲取系統环境資訊

Returns: str: JSON 格式的系統資訊

Inputs: no named parameters. Returns: Generic wrapper for non-object return types.

interactive_feedbackA

Interactive feedback collection tool for LLM agents.

USAGE RULES:

  1. During any process, task, or conversation, whether asking, replying, or completing phased tasks, you must call this tool to ask for feedback.

  2. Unless receiving termination instructions, all steps must repeatedly call this tool.

  3. Whenever user feedback is received, if the feedback content is not empty, you must call this tool again and adjust behavior based on the feedback content.

  4. Only when the user explicitly indicates "end" or "no more interaction needed" can you stop calling this tool, and the process is considered complete.

  5. You should summarize what have done, and provide project directory through args to let user know what you have done to provide feedback for next step.

IMPORTANT: The response may include a [NEW TASK] prefix. This means the user explicitly wants to switch to a completely new task. When you see [NEW TASK], you should: (1) deprioritize any ongoing work, (2) clear your TODO list, and (3) focus entirely on the new request that follows.

Args: project_directory: Project directory path for context summary: Summary of AI work completed for user review timeout: Timeout in seconds for waiting user feedback (default: 600 seconds)

Returns: list: List containing TextContent and MCPImage objects representing user feedback

Inputs: summary, timeout, project_directory.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Deploy Server
    One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

    TDQS

    A3.9/5.0

    Scored across 2 tools

    Disambiguation5/5

    The two tools serve completely different purposes—feedback collection and system information—so there is no risk of misselection. An agent would never confuse them.

    Naming Consistency3/5

    Both tools use snake_case, but the patterns differ: 'get_system_info' follows a verb_noun style, while 'interactive_feedback' is adjective_noun. This inconsistency makes the naming slightly less predictable.

    Tool Count3/5

    With only two tools, the set is minimal and borderline appropriate. The inclusion of an unrelated system utility alongside the core feedback tool dilutes the server's focus.

    Completeness4/5

    The interactive_feedback tool appears to cover the core feedback loop with its built-in usage rules, making the feedback process self-contained. However, there are no supporting tools for managing or analyzing feedback, and the system info tool doesn't contribute to the domain.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Resources

    Related MCP Connectors

    Related MCP Servers

    • MCP Feedback Enhanced

      Littlecowherd
      A
      license
      A
      quality
      D
      maintenance
      Establishes feedback-oriented development workflows with Web UI and desktop app, enabling AI to confirm with users before actions and consolidate tool calls into feedback requests.
      2
      13 PyPI
      MIT

    • mcp-feedback-collector

      sanshao85
      A
      license
      Not graded
      quality
      D
      maintenance
      A MCP server that enables AI assistants to collect interactive user feedback with text and image support via a modern GUI.
      234
      MIT

    • mcp-feedback-enhanced

      Minidoracat
      A
      license
      B
      quality
      B
      maintenance
      Establishes feedback-oriented development workflows with Web UI and desktop application dual interfaces, enabling AI to confirm with users and consolidate tool calls into feedback requests to reduce costs and improve development efficiency.
      2
      1,453 PyPI
      3,752
      MIT