Browser-use MCP Server
# browser-use MCP Server
### 🔥对应视频: https://youtu.be/jsd8TpzicRQ
### 🔥我的微信:stoeng
A Model Context Protocol server
This is a TypeScript-based MCP server that implements a simple notes system. It demonstrates core MCP concepts by providing:
- Resources representing text notes with URIs and metadata
- Tools for creating new notes
- Prompts for generating summaries of notes
## Features
### Resources
- List and access notes via `note://` URIs
- Each note has a title, content and metadata
- Plain text mime type for simple content access
### Tools
- `create_note` - Create new text notes
- Takes title and content as required parameters
- Stores note in server state
### Prompts
- `summarize_notes` - Generate a summary of all stored notes
- Includes all note contents as embedded resources
- Returns structured prompt for LLM summarization
## Development
Install dependencies:
```bash
npm install
```
Build the server:
```bash
npm run build
```
For development with auto-rebuild:
```bash
npm run watch
```
## Installation
To use with Claude Desktop, add the server config:
On MacOS: `~/Library/Application Support/Claude/claude_desktop_config.json`
On Windows: `%APPDATA%/Claude/claude_desktop_config.json`
```json
{
"mcpServers": {
"browser-use": {
"command": "/path/to/browser-use/build/index.js"
}
}
}
```
### Debugging
Since MCP servers communicate over stdio, debugging can be challenging. We recommend using the [MCP Inspector](https://github.com/modelcontextprotocol/inspector), which is available as a package script:
```bash
npm run inspector
```
The Inspector will provide a URL to access debugging tools in your browser.
TDQS
Scored across 1 tool
With only one tool, there is no possibility of ambiguity or overlap between tools. The single tool 'execute_browser_task' has a clearly distinct purpose that cannot be confused with any other tool in the set.
The naming is perfectly consistent as there is only one tool, following a clear verb_noun pattern ('execute_browser_task'). No other naming conventions are present to cause inconsistency.
A single tool for a browser automation server is too few for the apparent scope, as browser automation typically involves multiple operations like navigation, clicking, typing, or scraping. This feels thin and may limit functionality.
The tool surface is severely incomplete for browser automation, lacking essential operations such as navigating to URLs, interacting with elements, or extracting data. This will likely cause agent failures in handling common browser tasks.