Integration

Browse AI integration

Browse AI runs no-code web scraping robots that extract structured data (text fields and lists) from websites. An agent can list the robots of the account, run a robot on a URL and read the extracted data from the finished task, run a robot on many inputs at once (bulk runs), read the monitors that re-run a robot on a schedule, register and remove webhooks for task events, and check the platform status. Robots are trained, and monitors and robot cookies are created or changed, in the Browse AI dashboard; the internal teams route is not exposed. Authentication is an API key sent as a Bearer token.

Connect Browse AI How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in Browse AI.Built from the 14 Browse AI tools below, on the connection you add once.

What you can do with Browse AI

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

Bulk run, 3 operations

get_bulk_runGet bulk runGETGet one bulk run of a robot with the tasks run within it, including their captured data, paged.
get_bulk_runsGet bulk runsGETList the bulk runs of a robot, 10 per page. Pass a bulk run id to get_bulk_run for its tasks.
new_bulk_runNew bulk runPOSTRun a robot on many inputs at once, one task per entry of inputParameters (up to 50,000 tasks in one bulk run). Read the tasks and their captured data with get_bulk_run.

Task, 3 operations

get_robot_taskGet robot taskGETGet one task with its status and captured data (capturedTexts, capturedLists, capturedScreenshots). finishedAt stays null while the robot is still running; tasks time out after 15 minutes or the plan's maximum.
get_robot_tasksGet robot tasksGETList a robot's tasks (runs) with their status and captured data, 10 per page at most, filterable by status, bulk run and creation date range.
new_robot_taskNew robot taskPOSTRun a robot once (uses credits). Returns result.id at once with status in-progress: poll get_robot_task with that id until status is successful or failed, then read capturedTexts and capturedLists.

Webhook, 3 operations

create_new_webhookCreate new webhookPOSTRegister a URL that Browse AI calls when an event happens on a robot's tasks, for example taskFinished, so captured data can be received without polling.
delete_webhookDelete webhookDELETEDelete a webhook from a robot; its URL stops receiving events.
get_webhooksGet webhooksGETList the webhooks registered on a robot, with their URL and event type.

Monitor, 2 operations

get_monitorGet monitorGETGet one monitor of a robot: name, status (active or paused), schedule, input parameters and notification settings.
get_monitorsGet monitorsGETList the monitors of a robot: scheduled re-runs that detect changes in the captured data, with their schedule, input parameters and status.

Robot, 2 operations

get_robotGet robotGETGet one robot by id: its name, creation date and the input parameters a run can override (for example originUrl, the page the robot opens).
get_robotsGet robotsGETList the robots of the account with their ids and input parameters. Pass a robot id to new_robot_task, get_robot_tasks or get_monitors.

System status, 1 operation

get_system_statusGet system statusGETCheck whether the Browse AI task queue is operating (tasksQueueStatus OK) or under maintenance (UNDER_MAINTENANCE) before starting runs.

Use Browse AI from Claude Code, Cursor or Codex

Connect Browse AI in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its Browse AI operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect Browse AI

Create your workspace
Sign up free. Nothing to install.
Connect Browse AI once
Add your Browse AI connection in Settings. Every tool from this integration reuses it, and secrets are encrypted at rest and visible only to you.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.