Skip to main content
Glama

Agent Browser MCP Server

An MCP (Model Context Protocol) server that wraps agent-browser to provide browser automation tools for AI assistants. Supports both stdio and HTTP/SSE transports.

Features

40+ browser automation tools including:

Category

Tools

Navigation

browser_open, browser_navigate, browser_close

Interaction

browser_click, browser_fill, browser_type, browser_press, browser_hover, browser_select, browser_scroll

Inspection

browser_snapshot, browser_screenshot, browser_read, browser_get, browser_find

Execution

browser_eval, browser_batch, browser_wait

State

browser_cookie, browser_storage, browser_state, browser_session, browser_tab

DevTools

browser_console, browser_trace, browser_network, browser_react, browser_vitals, browser_a11y

Advanced

browser_frame, browser_stream, browser_dialog, browser_diff, browser_chat, browser_profile

System

browser_install, browser_doctor, browser_skills, browser_set

Related MCP server: selenium-mcp

Installation

npm install
npm run build

Usage

Stdio Mode (for local MCP clients)

npm run dev

HTTP/SSE Mode (for remote access)

node server.js

The server listens on port 5888 by default.

Environment Variables

Variable

Default

Description

PORT

5888

Server port

AGENT_BROWSER_SCREENSHOT_DIR

./screenshots

Directory for saved screenshots

BASE_URL

/screenshots

Public URL path for screenshots

MCP_SCRIPT_PATH

./dist/index.js

Path to compiled MCP server

ENABLE_HTTPS

false

Enable HTTPS with Let's Encrypt certs

CERT_PATH

/etc/letsencrypt/live/test2.infinitrocyber.com

SSL certificate path

Docker

docker build -t agentbrowser-mcp .
docker run -p 5888:5888 agentbrowser-mcp

Deployment

Render

The render.yaml blueprint is included. Deploy directly from the repo.

PM2

pm2 start ecosystem.config.cjs

API Endpoints

Method

Path

Description

GET

/health

Health check

GET

/mcp

SSE connection (creates new session)

POST

/mcp

JSON-RPC request

GET

/

Service info

Client Configuration

Claude Desktop

{
  "mcpServers": {
    "agent-browser": {
      "command": "node",
      "args": ["dist/index.js"]
    }
  }
}

Remote HTTP Client

{
  "mcpServers": {
    "agent-browser": {
      "url": "https://your-domain.com/mcp"
    }
  }
}

License

MIT

Available Tools

39 tools
browser_a11yD

Accessibility audit

Inputs: url, json, tags, session, +1 more.

browser_batchC

Execute multiple commands

Inputs: commands (required), bail.

browser_chatD

Natural language browser control

Inputs: instruction (required), session.

browser_clickC

Click element by selector or @ref

Inputs: selector (required), session.

browser_closeC

Close browser

Inputs: all, session.

browser_consoleC

View/clear console logs and errors

Inputs: action (required), json, session.

browser_dialogD

Handle dialogs

Inputs: action (required), text, session.

browser_diffC

Diff snapshots/screenshots

Inputs: type (required), json, target, session, +2 more.

browser_doctorD

Diagnose install

Inputs: fix, offline.

browser_evalD

Execute JavaScript

Inputs: script (required), session.

browser_fillC

Clear and fill input

Inputs: text (required), selector (required), session.

browser_findC

Find elements semantically

Inputs: type (required), value (required), action (required), name, +1 more.

browser_frameD

Switch frame

Inputs: session, selector.

browser_getC

Get text/html/value/url/title

Inputs: property (required), session, selector.

browser_hoverD

Hover element

Inputs: selector (required), session.

browser_installC

Install Chrome

Inputs: deps.

browser_navigateC

Back/forward/reload

Inputs: direction (required), session.

browser_networkD

Network requests/routes

Inputs: action (required), url, body, type, +7 more.

browser_openC

Open URL in browser

Inputs: url (required), session.

browser_pressC

Press key (Enter, Tab, etc)

Inputs: key (required), session.

browser_profileD

Chrome profiles

Inputs: action (required), name.

browser_reactD

React devtools

Inputs: action (required), json, fiberId, session.

browser_readC

Read page content or fetch URL

Inputs: url, filter, session.

browser_screenshotD

Take screenshot

Inputs: full, path, session.

browser_scrollC

Scroll page

Inputs: direction (required), pixels, session, selector.

browser_selectC

Select dropdown option

Inputs: value (required), selector (required), session.

browser_sessionD

Session management

Inputs: action (required), scope, prefix.

browser_setC

Set viewport/device/geo/media

Inputs: value (required), setting (required), session.

browser_skillsD

List/get skills

Inputs: action (required), all, name.

browser_snapshotC

Get accessibility tree with refs

Inputs: session.

browser_stateC

Save/load auth state

Inputs: action (required), path, session.

browser_storageC

Manage localStorage/sessionStorage

Inputs: type (required), action (required), key, value, +1 more.

browser_streamD

WebSocket streaming

Inputs: action (required), port, session.

browser_tabD

Manage tabs

Inputs: action (required), url, label, target, +1 more.

browser_traceC

Record trace

Inputs: action (required), path, session.

browser_typeC

Type text into input

Inputs: text (required), selector (required), session.

browser_vitalsD

Web Vitals metrics

Inputs: url, json, session.

browser_waitC

Wait for element/text/URL

Inputs: url, text, session, timeout, +1 more.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    C2.4/5.0

    Scored across 39 tools

    Disambiguation5/5

    Each tool targets a distinct browser operation (e.g., click, fill, hover, scroll) with no overlapping responsibilities. Even similar tools like get and read are differentiated by specific data returning or page fetching, and snapshot is distinct from screenshot.

    Naming Consistency5/5

    All tools follow a consistent 'browser_<verb>' pattern (e.g., open, click, fill, type, press), with no mixed conventions or vague names. The naming is predictable and clearly conveys each tool's action.

    Tool Count2/5

    With 39 tools, the count exceeds the typical range and feels heavy for most workflows. While the browser automation domain is broad, many tools are granular (e.g., browser_console, browser_trace), leading to potential selection overhead.

    Completeness4/5

    The tool set covers a comprehensive range of browser interactions: navigation, element actions, state management, debugging, performance monitoring, and network control. Minor gaps exist (e.g., file upload, form submission explicit handling), but most core workflows are well-supported.

    Maintenance

    ActivityInactive
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      D
      maintenance
      Enables AI agents to automate web tasks such as browsing, clicking, typing, and taking screenshots via the Model Context Protocol.
      1
      MIT
    • A
      license
      A
      quality
      D
      maintenance
      Enables AI-powered browser automation controlled through natural language, integrating Playwright with the Model Context Protocol to perform web interactions like navigation, form filling, and screenshots.
      10
      MIT