Server facts
BrowserTools MCP
Project
- Author
- AgentDeskAI
- Repository
- github.com/AgentDeskAI/browser-tools-mcp
- Website
- agentdesk.ai/
- License
- MIT
- Category
- Browser Automation
Tech
- Runtime / language
- node
- Transports
- stdio
- Remote available
- No
- Authentication
- Not specified
Activity
- GitHub stars
- 7,328
- Last commit
- Data as of
Tags
- Browser Debugging
- Chrome
- Lighthouse
- Web Audits
Overview
BrowserTools MCP gives AI coding clients visibility into a browser session through a Chrome extension and local Node services. The extension collects console output, network activity, screenshots, and the currently selected DOM element. A middleware server handles communication with the MCP server, processes captured information, and limits oversized or repeated log content. The README also describes removing cookies and sensitive headers before information reaches an LLM client. This makes the project relevant to developers investigating front-end errors without manually copying every piece of browser evidence into an editor.
Beyond inspection, the server exposes Lighthouse-based checks for accessibility, loading performance, search optimization, and general development practices. Dedicated modes sequence audit or debugging tools, and a NextJS-specific audit supplies guidance for that framework. Results include structured scores and issue details that compatible clients can interpret. Installation is not a single-process setup: users need the Chrome extension, the MCP package inside their client, and a separate browser-tools-server process. The BrowserTools panel must be opened in Chrome DevTools, and troubleshooting guidance recommends keeping only one such panel open. Most importantly, the README begins with a notice that development is inactive and asks users to choose another solution; the documented capabilities should therefore be evaluated with that maintenance limitation in mind.
Install & configure
Requirements
- Chrome with the BrowserTools extension
- A compatible MCP client
- Node.js tooling
- A separate browser-tools-server process
- The BrowserToolsMCP DevTools panel open
npx @agentdeskai/browser-tools-mcp@latestRun npx @agentdeskai/browser-tools-server@latest separately. The MCP package belongs in the client configuration. The project is no longer active.
Examples are translated from the published configuration. Supply your own credentials and verify requirements in the upstream README. Remote connectors are shown only where a supported configuration format is known.
Config file · macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
Use Settings → Developer → Edit Config when available. Do not assume a Linux configuration path; check the client documentation for your platform. Remote servers use the app’s connector flow when supported, not the local stdio file shown here. Merge this example with existing settings; do not overwrite unrelated servers.
{
"mcpServers": {
"browser-tools-mcp": {
"command": "npx",
"args": [
"@agentdeskai/browser-tools-mcp@latest"
],
"env": {}
}
}
}Config file · Project: .cursor/mcp.json
User: ~/.cursor/mcp.json
The project path is relative to the workspace root. The tilde denotes your user home directory. Use Cursor’s MCP settings to confirm which configuration and scope your installed version is reading. Merge this example with existing settings; do not overwrite unrelated servers.
{
"mcpServers": {
"browser-tools-mcp": {
"command": "npx",
"args": [
"@agentdeskai/browser-tools-mcp@latest"
],
"env": {}
}
}
}Config file · Workspace: .vscode/mcp.json
For user scope, use the “MCP: Open User Configuration” command if available. Do not guess a user configuration path across operating systems, profiles, containers, or remote workspaces; check the client documentation. Merge this example with existing settings; do not overwrite unrelated servers.
{
"servers": {
"browser-tools-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"@agentdeskai/browser-tools-mcp@latest"
],
"env": {}
}
}
}Config file · macOS/Linux: ~/.codeium/windsurf/mcp_config.json
Windows: %USERPROFILE%\.codeium\windsurf\mcp_config.json
Open the MCP configuration through Windsurf’s settings when possible to confirm the active file. Paths can differ in managed or future releases; check the client documentation if this location is not present. Merge this example with existing settings; do not overwrite unrelated servers.
{
"mcpServers": {
"browser-tools-mcp": {
"command": "npx",
"args": [
"@agentdeskai/browser-tools-mcp@latest"
],
"env": {}
}
}
}Config file · Project scope: .mcp.json at the repository root
The project file uses mcpServers. For user scope, use claude mcp add --scope user rather than assuming .mcp.json in your home directory is the user configuration. Check the client documentation for managed settings and storage details. Merge this example with existing settings; do not overwrite unrelated servers.
{
"mcpServers": {
"browser-tools-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"@agentdeskai/browser-tools-mcp@latest"
],
"env": {}
}
}
}Examples are translated from the published configuration. Supply your own credentials and verify requirements in the upstream README. Remote connectors are shown only where a supported configuration format is known.
Tools
Source: repository README. Tool descriptions are a documented snapshot, not a live capability check.
| Tool | What it does |
|---|---|
| runAccessibilityAudit | Check the page for accessibility problems, including WCAG-related issues. |
| runPerformanceAudit | Inspect page loading and rendering performance. |
| runSEOAudit | Review page-level search optimization factors. |
| runBestPracticesAudit | Check the page against web development practices. |
| runAuditMode | Run auditing tools in sequence, adding a NextJS check when detected. |
| runNextJSAudit | Provide NextJS-focused SEO and development guidance. |
| runDebuggerMode | Execute the debugging tools in their prescribed order. |
Names, permissions, and defaults can change. Check the current upstream documentation before granting access.
Use cases
- Investigate front-end failures using console and network evidence.
- Capture screenshots and inspect a selected DOM element.
- Run accessibility, performance, and SEO checks from an AI editor.
- Sequence auditing or debugging operations through dedicated modes.
Strengths
- Combines browser evidence with Lighthouse audit results.
- Documents local log storage and sensitive-header removal.
- Works with multiple MCP clients, with particular attention to Cursor.
Limitations
- The project explicitly states that it is no longer active.
- Requires an extension and two distinct server components.
Frequently asked questions
Independent listing, not an affiliation or security review. Grant the least access needed and verify the upstream instructions before running code.
