loaditout.ai
SkillsPacksTrendingLeaderboardAPI DocsBlogSubmitRequestsCompareAgentsXPrivacyDisclaimer
{}loaditout.ai
Skills & MCPPacksBlog

babelwrap-js

MCP Tool

babelwrap/babelwrap-js

JavaScript/TypeScript SDK for BabelWrap — the web, as an API, for your agents.

Install

$ npx loaditout add babelwrap/babelwrap-js

Platform-specific configuration:

.claude/settings.json
{
  "mcpServers": {
    "babelwrap-js": {
      "command": "npx",
      "args": [
        "-y",
        "babelwrap-js"
      ]
    }
  }
}

Add the config above to .claude/settings.json under the mcpServers key.

About

BabelWrap JavaScript/TypeScript SDK

The official JavaScript/TypeScript SDK for BabelWrap — the web, as an API, for your agents.

Installation
npm install babelwrap
Quick Start
import { BabelWrap } from "babelwrap";

const bw = new BabelWrap({ apiKey: process.env.BABELWRAP_API_KEY! });
const session = await bw.createSession();

try {
  // Navigate to a page
  const snap = await session.navigate("https://news.ycombinator.com");
  console.log(snap.title); // "Hacker News"

  // Extract structured data
  const stories = await session.extract("top 5 story titles and their URLs");
  console.log(stories);
} finally {
  await session.close();
}
With TypeScript using (auto-cleanup)
import { BabelWrap } from "babelwrap";

const bw = new BabelWrap({ apiKey: process.env.BABELWRAP_API_KEY! });
await using session = await bw.createSession();

await session.navigate("https://example.com");
await session.fill("Email field", "user@example.com");
await session.fill("Password field", "secret");
await session.submit();

const data = await session.extract("the user profile information");
API Reference
BabelWrap
const bw = new BabelWrap({
  apiKey: "bw_...",          // Required
  baseUrl: "https://...",    // Default: https://api.babelwrap.com
  timeout: 60000,            // Default: 60s
  maxRetries: 3,             // Default: 3
});

Methods:

  • createSession(options?) — Create a new browser session
  • usage() — Get current usage stats
  • health() — Health check
  • mapSite(url, cookies?) — Map a website and generate typed tools
  • listSites() — List mapped sites
  • siteTools(siteId) — Get tools for a mapped site
  • executeTool(siteId, toolName, params?) — Execute a site tool
Session

All methods return a Snapshot (attribute-accessible page state) unless noted.

  • navigate(url) — Navigate to URL
  • click(target) — Click element by description

-

Tags

ai-agentsbrowser-automationjavascriptmcpsdktypescriptweb-scraping

Reviews

Loading reviews...

Quality Signals

0
Installs
Last updated5 days ago
Security: BREADME
New

Safety

Risk Levelmedium
Data Access
read
Network Accessnone

Details

Sourcegithub-crawl
Last commit4/7/2026
View on GitHub→

Embed Badge

[![Loaditout](https://loaditout.ai/api/badge/babelwrap/babelwrap-js)](https://loaditout.ai/skills/babelwrap/babelwrap-js)