> ## Documentation Index
> Fetch the complete documentation index at: https://docs.agentruntime.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Browserless connector

> Capture screenshots, generate PDFs, scrape pages, and run Lighthouse audits via Browserless headless Chrome.

The **browserless** MCP adapter connects to [Browserless](https://www.browserless.io/) for headless browser automation. Render pages to images or PDFs, extract structured data with CSS selectors, run BrowserQL queries, and audit site performance — without hosting your own Chrome fleet.

## Prerequisites

* A Browserless account with an API token (cloud or self-hosted)
* For self-hosted Browserless: the REST base URL for your deployment
* **project\_contributor** access in AgentRuntime

## Connect in AgentRuntime

<Steps>
  <Step title="Get a Browserless API token">
    Sign in to the [Browserless dashboard](https://www.browserless.io/) and copy your API token. Note your region or self-hosted base URL if applicable.
  </Step>

  <Step title="Create a connection">
    Go to **Connections** and click **New custom connection** to create an API key connection with `APIToken`. Set `APIBase` if you use a custom or self-hosted endpoint.
  </Step>

  <Step title="Add browserless MCP instance">
    Go to **MCP**, click **Add instance** for **browserless**, wire your connection on **Instance config**, set the profile **active**, and save.
  </Step>

  <Step title="Smoke test">
    Run **mcp\_call** → `capture_screenshot` with a public URL to confirm rendering works.
  </Step>
</Steps>

See [Authentication](/integrations/authentication) for binding connections to MCP instances.

## What you can build

* **Competitive snapshots** — `capture_screenshot` of competitor pricing pages → store in S3 or attach to a Wrike task.
* **Report PDFs** — LLM generates HTML summary → `generate_pdf` → email or upload via another connector.
* **Price monitoring** — `scrape_url` with CSS selectors on product pages → compare to Postgres baseline → alert on change.
* **Performance gates** — `get_website_performance` after deploy → fail workflow if Lighthouse score drops below threshold.
* **Dynamic scraping** — `run_bql_query` for multi-step navigation before extraction.

## Tools

| Tool                      | Description                                                |
| ------------------------- | ---------------------------------------------------------- |
| `capture_screenshot`      | Render a URL or inline HTML to PNG, JPEG, or WebP (base64) |
| `generate_pdf`            | Render a URL or inline HTML to a PDF document              |
| `scrape_url`              | Load a page and extract fields via CSS selectors           |
| `run_bql_query`           | Execute a BrowserQL GraphQL query or mutation              |
| `get_website_performance` | Lighthouse-backed performance audit for a URL              |

## Example

Capture a screenshot after a deploy verification step:

```json theme={null}
{
  "id": "screenshot-home",
  "type": "mcp_call",
  "name": "Screenshot homepage",
  "tool_name": "capture_screenshot",
  "tool_args": {
    "url": "https://example.com",
    "options": {
      "fullPage": true,
      "type": "png"
    }
  },
  "timeout_s": 60
}
```

## Configuration

| Key        | Required | Description                                            |
| ---------- | -------- | ------------------------------------------------------ |
| `APIToken` | Yes      | Browserless API token (sent as query param)            |
| `APIBase`  | No       | REST base URL without path (default Browserless cloud) |

## Troubleshooting

| Issue                | Fix                                                                                    |
| -------------------- | -------------------------------------------------------------------------------------- |
| Timeout on render    | Increase workflow `timeout_s`; complex SPAs may need longer Browserless session limits |
| Empty scrape results | Page may require JS; verify selectors against rendered DOM, not static HTML            |
| `401` / quota errors | Check token, plan limits, and concurrent session caps in Browserless dashboard         |
| PDF layout broken    | Pass print-specific CSS in `html` or tune `options` for page size and margins          |

## Related

* [Integrations quickstart](/integrations/quickstart)
* [Connector catalog](/integrations/connector-catalog)
* [Troubleshooting](/platform/troubleshooting)