Supports launching and controlling Firefox browser sessions for web automation, including navigation, element interaction, screenshots, and state management.
Provides Selenium browser automation tools, including session management, navigation, interaction, snapshots, screenshots, tab management, and network mocking via WebDriver.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@se-mcp Open https://news.ycombinator.com, get the top 5 story titles, and return them as a list"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
se-mcp is the recommended way for MCP-aware AI clients (VS Code, Claude Desktop, Cursor, etc.) to use se-cli — a token-efficient Selenium browser automation CLI.
This package is a thin wrapper around se-cli's built-in MCP server. It provides a standalone entry point (npx @browsers-cli/se-mcp) so MCP clients can discover and invoke all browser automation tools without manually configuring the se-cli binary.
┌──────────────────┐ JSON-RPC 2.0 ┌──────────────────┐
│ MCP Client │ ◄─── over stdio ───► │ se-mcp │
│ (VS Code, │ │ (this package) │
│ Claude, etc.) │ │ thin wrapper │
└──────────────────┘ └────────┬─────────┘
│
require() │ delegates to
▼
┌──────────────────┐
│ @browsers-cli/ │
│ se-cli │
│ MCP Server │
│ (mcp-server.ts) │
└────────┬─────────┘
│
JSON line │ over socket
▼
┌──────────────────┐
│ se-cli Daemon │
│ (long-lived) │
│ WebDriver + │
│ Selenium │
└────────┬─────────┘
│
▼
┌──────────────────┐
│ Browser │
│ (Chrome/Edge/ │
│ Firefox) │
└──────────────────┘The MCP server reads JSON-RPC 2.0 requests from stdin and writes responses to stdout. Browser commands are forwarded to the se-cli daemon over a Unix socket (or Windows named pipe), which holds the WebDriver instance across calls.
Related MCP server: Selenium MCP Server
You do not need to install anything globally. MCP clients will use npx automatically:
npx @browsers-cli/se-mcpnpm install -g @browsers-cli/se-mcpNote: se-mcp depends on
@browsers-cli/se-clias a peer dependency. It will be installed automatically bynpxor when you install se-mcp with npm 7+.
Add to your VS Code settings (settings.json):
{
"mcp.servers": {
"se-cli": {
"command": "npx",
"args": ["-y", "@browsers-cli/se-mcp"]
}
}
}Or use the bundled server.json file directly — copy its contents into your MCP client configuration.
Add to claude_desktop_config.json:
{
"mcpServers": {
"se-cli": {
"command": "npx",
"args": ["-y", "@browsers-cli/se-mcp"]
}
}
}Add to ~/.cursor/mcp.json:
{
"mcpServers": {
"se-cli": {
"command": "npx",
"args": ["-y", "@browsers-cli/se-mcp"]
}
}
}Add to .mcp.json in your project root:
{
"mcpServers": {
"se-cli": {
"command": "npx",
"args": ["-y", "@browsers-cli/se-mcp"]
}
}
}Use the server.json file included in this package:
{
"mcpServers": {
"se-cli": {
"command": "npx",
"args": ["-y", "@browsers-cli/se-mcp"],
"env": {}
}
}
}se-mcp exposes 50+ MCP tools covering the full browser automation lifecycle. All tools are prefixed with browser_.
Tool | Description |
| Start a browser session (Chrome, Edge, or Firefox) |
| Close a browser session and stop the daemon |
| List all active browser sessions |
| Close all active browser sessions |
Tool | Description |
| Navigate to a URL |
| Navigate back in history |
| Navigate forward in history |
| Reload the current page |
| Get the current page title |
| Get the current page URL |
Tool | Description |
| Click an element by ref or CSS selector |
| Fill an input field with text |
| Type text into the focused element |
| Press a keyboard key |
| Select a dropdown option |
| Check a checkbox |
| Uncheck a checkbox |
| Hover over an element |
| Double-click an element |
| Drag and drop between elements |
Tool | Description |
| Take an accessibility (aria) snapshot with element refs |
| Find elements by text content |
| Take a screenshot of the page or element |
| Execute JavaScript in the page |
Tool | Description |
| List all open tabs |
| Open a new tab |
| Close the current tab |
| Switch to a tab by index |
Tool | Description |
| List all cookies |
| Get a specific cookie |
| Set a cookie |
| Delete a cookie or all cookies |
| Save browser state to JSON |
| Load browser state from JSON |
Tool | Description |
| Accept an alert/confirm/prompt dialog |
| Dismiss a dialog |
| Upload a file to a file input |
| Set the viewport size |
| Press and hold a key |
| Release a held key |
| Move mouse to coordinates |
| Press a mouse button |
| Release a mouse button |
| Scroll the mouse wheel |
| Chain multiple actions in one call |
Tool | Description |
| Assert conditions (visible, hidden, text, value, etc.) |
Tool | Description |
| Outline elements for visual debugging |
| Get buffered console messages and JS errors |
| List buffered network requests |
| Show details of a specific request |
| Mock a network route |
| List active route mocks |
| Remove a route mock |
MCP client (e.g., VS Code) spawns npx @browsers-cli/se-mcp as a subprocess.
se-mcp loads the MCP server from @browsers-cli/se-cli and starts listening on stdin/stdout.
When the MCP client calls a tool (e.g., browser_open), the server maps it to se-cli CLI args and sends them to the se-cli daemon over a socket.
The daemon holds the WebDriver instance and executes the command, returning a JSON response.
The server wraps the response in a JSON-RPC 2.0 envelope and writes it to stdout.
The daemon architecture means the browser stays open across tool calls. There is no reconnection cost, no driver restart overhead, and the WebDriver instance is reused. This is what makes se-cli token-efficient compared to traditional MCP browser implementations.
The browser_snapshot tool returns a compact YAML-like accessibility tree with element references (e1, e2, etc.). Subsequent commands can reference these elements by ref instead of verbose CSS selectors, dramatically reducing token usage.
You can also use se-mcp programmatically:
const { startMcpServer, McpServer, toolDefinitions, mapToolToCliArgs } = require('@browsers-cli/se-mcp');
// Start the MCP server (reads JSON-RPC from stdin, writes to stdout)
startMcpServer();
// Or create an instance with a custom workspace directory
const server = new McpServer('/path/to/workspace');
server.start();import { startMcpServer, McpServer, toolDefinitions, mapToolToCliArgs } from '@browsers-cli/se-mcp';
// startMcpServer(workspaceDir?: string): void
// McpServer: class { constructor(workspaceDir?: string); start(): void }
// toolDefinitions: ToolDef[]
// mapToolToCliArgs(toolName: string, args: any): string[] | nullse-cli — The core Selenium CLI with daemon architecture
se-cli documentation — Full documentation and guides
se-cli npm — Core package on npm
Apache License 2.0 — see LICENSE for details.
This server cannot be installed
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
MCP server exposing the Backtest360 engine API as tools for AI agents.
MCP server for AI agents to plan, verify, and deploy Cloudflare-native apps.
A comprehensive Model Context Protocol (MCP) server that enables AI assistants to interact with yo…
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/se-cli/se-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server