@browsa/mcp

Drive anti-detect browsers from Claude Desktop, Cursor, Cline, and any MCP-compatible client.

MCP (Model Context Protocol) server for Browsa — turn any LLM into a real-browser agent with one config line. The agent gets a freshly-spawned macOS-Chrome-fingerprinted Chromium with built-in CAPTCHA solving, human-shaped behavioral pacing, and a live noVNC stream so you can watch it work.

What this is

Most "AI agent" setups stop at the LLM. We give the LLM a real browser. Through MCP, Claude can now:

• Navigate, fill forms, click, scroll, scrape
• Pass through Cloudflare Turnstile, hCaptcha, reCAPTCHA, ALTCHA
• Run from a residential US/UK/DE/etc. exit IP
• Stay logged in across calls via persistent identities

No code from you — just configure Claude Desktop (or Cursor / Cline / Zed) once, and ask in natural language.

Install

Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "browsa": {
      "command": "npx",
      "args": ["-y", "@browsa/mcp"],
      "env": {
        "BROWSA_API_KEY": "agt_live_...",
        "BROWSA_LLM_API_KEY": "sk-ant-..."
      }
    }
  }
}

Restart Claude Desktop. Ask: "Go to news.ycombinator.com and tell me the top story title."

Cursor

~/.cursor/mcp.json:

{
  "mcpServers": {
    "browsa": {
      "command": "npx",
      "args": ["-y", "@browsa/mcp"],
      "env": { "BROWSA_API_KEY": "agt_live_...", "BROWSA_LLM_API_KEY": "sk-ant-..." }
    }
  }
}

Cline / Continue / Zed

Any MCP-spec-compliant client. Point at npx -y @browsa/mcp and set the two env vars.

Required env

| Var | Required | What | |---|---|---| | BROWSA_API_KEY | yes | Your agt_live_* key from browsa.io → Keys | | BROWSA_LLM_API_KEY | recommended | Default LLM provider key (e.g. Anthropic). Without this, Claude must pass it on every run_task call | | BROWSA_LLM | no | Default LLM. Defaults to claude-opus-4-7 | | BROWSA_COUNTRY | no | Default egress country (e.g. US) | | BROWSA_BASE_URL | no | Override prod URL (staging / self-hosted) |

Tools exposed

| Tool | What | |---|---| | run_task | Run a browser-driven task. Returns final result + live noVNC URL. | | get_job | Poll a long-running task. Optional server-side wait_seconds. | | cancel_job | Cancel a running task. | | respond_to_job | Answer an input_required prompt (CAPTCHA, decision). | | credits | Show credit balance + top-up packs. | | list_webhooks / create_webhook / delete_webhook | Webhook CRUD. | | list_identities | List persistent browser identities (use one via profile_id to reuse warmed cookies). |

Why MCP-first matters

Most agent platforms gate distribution behind "write code → deploy → call our API." We gate it behind "paste config → ask Claude." For an AI agent platform, the prospect's first 30 seconds is the make-or-break moment. With this MCP server, the first 30 seconds is a working browser doing what they asked.

Source + Issues

• Source: https://github.com/mohasaaid/neout/tree/main/sdks/browsa-mcp
• Issues: https://github.com/mohasaaid/neout/issues
• API reference: https://browsa.io/docs

License

MIT.