Web Browser

Capabilities

• • Open a page at a given URL in a headless browser
• • Extract page content to structured Markdown text
• • Click an element identified by a selector or by visible text
• • Fill and submit a form with a set of field values
• • Scroll the page and wait until a specified selector is visible
• • Route a request through the assigned proxy address from the pool
• • Inspect response status and response headers of a page
• • Return a clear error when a page is blocked or times out

Constraints

• • Cap the work at 60 seconds per page. Return a timeout error.
• • Do not submit a payment or change account data without an explicit user instruction.
• • Do not bypass an access control of the target site. Report the block.
• • Do not keep cookies or page data beyond the current request.

MCP schema

{
  "name": "web_browser",
  "input": {
    "type": "object",
    "required": [
      "action",
      "url"
    ],
    "properties": {
      "url": {
        "type": "string"
      },
      "action": {
        "enum": [
          "open",
          "extract",
          "click",
          "fill_submit",
          "wait_for"
        ]
      },
      "values": {
        "type": "object"
      },
      "selector": {
        "type": "string"
      },
      "visible_text": {
        "type": "string"
      },
      "timeout_seconds": {
        "type": "integer"
      }
    }
  },
  "output": {
    "type": "object",
    "properties": {
      "markdown": {
        "type": "string"
      },
      "final_url": {
        "type": "string"
      },
      "duration_ms": {
        "type": "integer"
      },
      "status_code": {
        "type": "integer"
      }
    }
  },
  "description": "Opens a page in a headless browser, waits for the target, and returns the extracted content."
}