Integration

BrowserAct integration

BrowserAct runs browser bots (no-code scraping workflows and AI agent bots built in the BrowserAct dashboard) and official bot templates in the cloud, and returns the structured data they extract. List and inspect your published bots and the official templates with their input schema, start a run that waits for the extracted output, follow, resume or cancel runs, and list past runs and the proxy regions a template run can use. Every run spends BrowserAct credits (about 3 credits per workflow step, rotating proxy traffic 5,000 credits per GB, per the pricing page); the free trial gives 2,000 credits and runs one task at a time. Building or editing bots, webhook callbacks, live browser view links, log links, screenshots and result file download links are not exposed: the extracted data comes back inline (up to 5 MB; a larger result is not returned here). A run visits whatever site its input names (many templates take any url or keywords) and its bot may solve CAPTCHAs or pause for a human step, so the caller chooses the target and answers for it. BrowserAct's terms require respecting each target site's terms of use and robots.txt. Authentication is a BrowserAct API key (Integrations, Manage API Keys) sent as a Bearer token.

Connect BrowserAct How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in BrowserAct.Built from the 12 BrowserAct tools below, on the connection you add once.

What you can do with BrowserAct

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.

Run, 4 operations

get_runGet runGETReads one bot run: status, input, failure details, credits spent, the steps it took and, once it has ended, its extracted output (inline, up to 5 MB). For status alone, get_run_status is lighter.
list_runsList runsGETLists bot runs (tasks), page by page (page, limit up to 100; total_pages and total_count in the answer). Filter by status, bot_id, a bot name keyword, or an inclusive creation time range. Pass a task_id to get_run.
run_botRun botPOSTRuns a published bot for real in a cloud browser and returns its extracted output. Spends BrowserAct credits. Input follows get_bot's input_schema. Waits for the job and returns the finished result; gives up after 10 minutes.
run_templateRun templatePOSTRuns an official template for real in a cloud browser and returns its extracted output. Spends credits. Input follows get_template's input_schema. Waits for the job and returns the finished result; gives up after 10 minutes.

Bot, 2 operations

get_botGet botGETReads one bot: name, type, creation and publication time, and the input_schema (JSON Schema) its run input must follow, plus its output_schema. An unpublished bot has null schemas and cannot be run (error 10020).
list_botsList botsGETLists your bots that have a runnable published version, page by page (page, limit up to 100; total_pages and total_count in the answer). Filter by bot_type (workflow or agent) or a name keyword. Pass a bot_id to get_bot or run_bot.

Template, 2 operations

get_templateGet templateGETReads one official bot template: name, description, publication time and the input_schema (JSON Schema) that run_template's input must follow.
list_templatesList templatesGETLists the official bot templates curated by BrowserAct, page by page (page, limit up to 500; total_pages and total_count in the answer), with an optional name keyword. Pass a template_id to get_template or run_template.

Cancel, 1 operation

cancel_runCancel runPOSTCancels a running or paused run for real; a canceled run cannot be resumed. A run already finished, canceled or failed returns error 10121. Returns the run status.

Region, 1 operation

list_regionsList regionsGETLists the proxy regions your API key can use: a code and a readable name for each. Pass a code as proxy_region to run_template.

Resume, 1 operation

resume_runResume runPOSTAsks BrowserAct to resume a paused run for real, so the bot continues working. Resume is asynchronous: the returned status can still read paused. Not paused: error 10127; agent bot runs cannot resume (10879).

Status, 1 operation

get_run_statusGet run statusGETReads only the current status of a run (created, running, pausing, paused, finished, canceled or failed) and when it last changed. Lighter than get_run; use it to follow a run.

Use BrowserAct from Claude Code, Cursor or Codex

Connect BrowserAct in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its BrowserAct 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 BrowserAct

Create your workspace
Sign up free. Nothing to install.
Connect BrowserAct once
Add your BrowserAct 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.