Skip to main content
Glama
io.github.soulfir/babelwrap-mcp

io.github.soulfir/babelwrap-mcp

by babelwrap

BabelWrap MCP Server

An MCP server that gives AI agents web browsing superpowers via the BabelWrap API.

Works with Claude Desktop, Cursor, Claude Code, and any MCP-compatible client.

Installation

# Using uvx (recommended -- no install required)
uvx babelwrap-mcp

# Using pip
pip install babelwrap-mcp

# Using pipx (isolated environment)
pipx install babelwrap-mcp

Related MCP server: Cloudflare Playwright MCP

Setup

1. Get an API Key

Sign up at babelwrap.com and create an API key from your dashboard.

2. Configure Your MCP Client

Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "babelwrap": {
      "command": "uvx",
      "args": ["babelwrap-mcp"],
      "env": {
        "BABELWRAP_API_KEY": "bw_your_api_key_here"
      }
    }
  }
}

If you installed with pip or pipx, use the binary directly:

{
  "mcpServers": {
    "babelwrap": {
      "command": "babelwrap-mcp",
      "args": [],
      "env": {
        "BABELWRAP_API_KEY": "bw_your_api_key_here"
      }
    }
  }
}

Claude Code

claude mcp add babelwrap -- uvx babelwrap-mcp

Then set your API key as an environment variable:

export BABELWRAP_API_KEY="bw_your_api_key_here"

Available Tools

Tool

Description

babelwrap_new_session

Create a new browser session

babelwrap_close_session

Close a browser session

babelwrap_navigate

Navigate to a URL

babelwrap_snapshot

Get current page state

babelwrap_click

Click an element

babelwrap_fill

Fill a form field

babelwrap_submit

Submit a form

babelwrap_extract

Extract structured data

babelwrap_screenshot

Take a screenshot

babelwrap_press

Press a keyboard key

babelwrap_scroll

Scroll the page

babelwrap_hover

Hover over an element

babelwrap_upload

Upload a file

babelwrap_back / babelwrap_forward

Browser history

babelwrap_wait_for

Wait for a condition

babelwrap_list_pages / babelwrap_switch_page

Multi-tab support

Environment Variables

Variable

Required

Description

BABELWRAP_API_KEY

Yes

Your BabelWrap API key

BABELWRAP_API_URL

No

API base URL (default: https://api.babelwrap.com/v1)

Documentation

Full documentation at babelwrap.com/docs/mcp

License

MIT

Available Tools

18 tools
babelwrap_backC

Navigate back in the browser history.

Inputs: session_id (required), compact. Returns: object

babelwrap_clickC

Click an element on the page using a natural language description.

Inputs: session_id (required), target (required), compact. Returns: object

babelwrap_close_sessionB

Close and clean up a browser session.

Inputs: session_id (required). Returns: object

babelwrap_extractB

Extract structured data from the current page using a natural language query.

Inputs: session_id (required), query (required), compact. Returns: object

babelwrap_fillB

Fill an input field with a value.

Inputs: session_id (required), target (required), value (required), compact. Returns: object

babelwrap_forwardA

Navigate forward in the browser history.

Inputs: session_id (required), compact. Returns: object

babelwrap_hoverC

Hover over an element on the page.

Inputs: session_id (required), target (required), compact. Returns: object

babelwrap_list_pagesA

List all open pages (tabs/popups) in this browser session.

Use this to discover tabs opened by popups or target="_blank" links, then switch to them with babelwrap_switch_page.

Inputs: session_id (required). Returns: object

babelwrap_navigateB

Navigate to a URL and get a structured snapshot of the page.

Inputs: session_id (required), url (required), compact. Returns: object

babelwrap_new_sessionB

Create a new browser session for web interaction.

Inputs: metadata. Returns: object

babelwrap_pressA

Press a keyboard key on the current page.

Inputs: session_id (required), key (required), compact. Returns: object

babelwrap_screenshotB

Take a screenshot of the current page (for debugging).

Inputs: session_id (required), compact. Returns: object

babelwrap_scrollC

Scroll the page up or down.

Inputs: session_id (required), direction, amount, compact. Returns: object

babelwrap_snapshotA

Get the current page state as a structured snapshot without taking any action.

Inputs: session_id (required), compact. Returns: object

babelwrap_submitC

Submit a form on the current page.

Inputs: session_id (required), target, compact. Returns: object

babelwrap_switch_pageA

Switch to a different page (tab/popup) within this session.

Use babelwrap_list_pages first to see available pages and their indices.

Inputs: session_id (required), index (required), compact. Returns: object

babelwrap_uploadB

Upload a file to a file input element on the page.

Inputs: session_id (required), target (required), file_base64 (required), filename (required), +1 more. Returns: object

babelwrap_wait_forA

Wait for a condition on the page before proceeding.

Useful after clicking a button that triggers async loading, SPA navigation, or any action where you need to wait for specific content to appear.

Inputs: session_id (required), text, selector, url_contains, +2 more. Returns: object

Tool Schema Changelog

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

Loading tool schema changelog…

    TDQS

    A3.5/5.0

    Scored across 18 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose: navigation, interaction, session management, page state, extraction, and waiting. No two tools overlap in functionality, and descriptions effectively differentiate them.

    Naming Consistency5/5

    All tools consistently use the 'babelwrap_' prefix with a verb_noun pattern in snake_case (e.g., babelwrap_click, babelwrap_new_session). The naming is uniform and predictable.

    Tool Count5/5

    With 18 tools, the server covers a comprehensive range of browser automation tasks (navigation, interaction, session management, page analysis) without being excessive or insufficient. The count is well-scoped for the domain.

    Completeness4/5

    The tool set covers core browser actions like navigation, clicking, filling, submitting, waiting, screenshots, and file uploads. Minor gaps exist (e.g., no JavaScript execution, cookie management, or advanced element queries), but the existing tools enable most workflows.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      Not graded
      maintenance
      Enables AI assistants to control a browser through a set of tools, allowing them to perform web automation tasks like navigation, typing, clicking, and taking screenshots.
      Apache 2.0
    • A
      license
      Not graded
      quality
      Not graded
      maintenance
      Enables AI assistants to control a browser through a set of tools, allowing them to perform web automation tasks like navigation, typing, clicking, and taking screenshots.
      Apache 2.0
    • A
      license
      Not graded
      quality
      Not graded
      maintenance
      Enables AI assistants to control a browser through a set of tools, allowing them to perform web automation tasks like navigation, typing, clicking, and taking screenshots.
      Apache 2.0