MCP Feedback Enhanced
# MCP Shouji
## 🔧 本地定制版本说明
> 感谢[f](https://github.com/f/mcp-shouji)的反馈项目,本版本修复了以下问题并新增功能:
### ✅ 已修复问题
1. **超时设置问题**:修复了超时时间永远是 600s 的问题,现已修复默认 24 小时超时
,可以长时间等待反馈啦
2. **图片上传问题**:修复了无法上传图片报错序列号错误的问题,现已支持更多类型图
片 😁🎉
3. **断网重连功能**:新增断网不断链接功能,现在可以离线同一个会话等待(适合使用
手机热点为电脑提供互联网的场景)
### 🚀 如何使用
Fork 本项目到本地,在 Cursor 中配置:
```json
{
"mcpServers": {
"mcp-shouji-local": {
"command": "uv",
"args": [
"run",
"--directory",
"C:\\Users\\Administrator\\Desktop\\project\\mcp-shouji-main",
"python",
"-m",
"mcp_shouji"
],
"timeout": 86400,
"env": {
"MCP_DEBUG": "false",
"MCP_WEB_HOST": "127.0.0.1",
"MCP_WEB_PORT": "8765",
"MCP_DESKTOP_MODE": "false",
"MCP_LANGUAGE": "zh-CN"
},
"autoApprove": ["interactive_feedback"]
}
}
}
```
> **注意**:请将
> `"C:\\Users\\Administrator\\Desktop\\project\\mcp-shouji-main"` 改
> 为您本地的项目位置,这样可以实现本地高自由度的定制。
### 📸 界面预览
<div align="center">
<img src="images/0.png" width="600" alt="MCP Shouji 界面预览" />
<br>
<em>主界面 - 支持提示管理、自动提交、会话跟踪上传图片等功能</em>
</div>
<div align="center">
<img src="images/1.png" width="600" alt="MCP Shouji 功能展示" />
<br>
<em>超时功能展示 - 智能工作流程和现代化体验</em>
</div>
---
TDQS
Scored across 2 tools
The two tools have clearly distinct purposes: get_system_info retrieves system environment information, while shouji provides an interactive web interface for collecting user input and attachments. There is no overlap in functionality, making it easy for an agent to choose the correct tool for each task.
The naming is inconsistent: get_system_info follows a clear verb_noun pattern, but shouji uses a non-English term (Chinese for 'collect') without a descriptive verb, breaking the pattern. This mix of conventions could confuse agents expecting a uniform naming style.
With only 2 tools, the server feels thin for a general-purpose feedback system, as it lacks tools for processing, analyzing, or managing feedback beyond collection. However, the count is borderline appropriate if the scope is limited to basic input gathering and system info retrieval.
For a feedback-enhanced server, there are significant gaps: it provides input collection and system info but lacks tools for storing, retrieving, analyzing, or responding to feedback. This incomplete surface will likely cause agent failures when trying to handle feedback workflows beyond simple input capture.