Feedback Collector MCP
# Feedback Collector MCP Tool
Collect user feedback with text and image support through an Electron app.
## Quick Setup
1. **Clone repository**:
```bash
git clone https://github.com/ceciliomichael/feedback-js.git
cd feedbackjs-mcp
```
2. **Install dependencies**:
```bash
npm install
```
3. **Build the application**:
```bash
npm run make
```
## Cross-Platform Building
The app can be built for Windows, macOS, and Linux platforms:
```bash
# Build for the current platform only
npm run make
# Build for specific platforms
npm run make:win # Windows only
npm run make:mac # macOS only
npm run make:linux # Linux only
# Build for all platforms
npm run make:all
```
> **Note:** Building for macOS from Windows or Linux requires additional setup:
> - For code signing, you need an Apple Developer account
> - For proper macOS builds from non-macOS platforms, consider using a CI service like GitHub Actions
## MCP Configuration
Add to your AI tool configuration (works with Claude Desktop, Cursor, and other MCP clients):
```json
{
"mcpServers": {
"feedback-collector": {
"command": "node",
"args": ["/absolute/path/to/mcp-server.js"]
}
}
}
```
## Tool Parameters
| Parameter | Type | Description | Default |
|-----------|------|-------------|---------|
| `prompt` | string | Message to display | "Please provide feedback" |
| `title` | string | Window title | "AI Feedback Collection" |
| `time_format` | enum | Time format (full, iso, date, time, unix) | "full" |
| `timezone` | string | Timezone | Local timezone |
## Features
- Text feedback with markdown prompt support
- Image uploads (file selection, drag-and-drop, clipboard paste)
- Quick response buttons (Submit, Approve, Enough, Cancel)
- Detailed time information TDQS
Scored across 1 tool
With only one tool, there is no possibility of confusion or overlap between tools. The single tool 'collect_feedback' has a clearly distinct purpose that cannot be mistaken for any other tool in this server.
The single tool name 'collect_feedback' follows a clear verb_noun pattern. With only one tool, there is no inconsistency in naming conventions to evaluate, making it perfectly consistent.
A single tool is too few for most server purposes, as it limits functionality and suggests a thin or incomplete implementation. While it might be appropriate for a very narrow scope, it generally feels insufficient for robust agent interaction.
The server appears to be for collecting feedback, but with only a collection tool, there are severe gaps: no tools for viewing, analyzing, updating, or managing feedback. This makes the surface severely incomplete for any meaningful feedback management workflow.