Browse AI: Actions and Triggers in CrocoClick Automations

Browse AI is a no-code web scraping and monitoring platform built around Robots: trained extractors that retrieve structured data from a web page. Each time a robot runs, it’s called a “Task,” and a batch run queues multiple tasks from the same robot all at once. Native integration brings these elements directly into the CrocoClick automation builder: your workflows can react to a completed scraping session and control Browse AI in the opposite direction by launching robots on demand to retrieve the captured data.

What the integration enables

The integration consists of two parts:

• Trigger (Browse AI to your automations): an instant trigger, “New Task Completed,” which fires as soon as a Browse AI robot finishes a task. Filterable by robot and selection operator.

• Actions (your automations to Browse AI): four actions covering single or batch execution, plus retrieving results by ID.

Connect Browse AI

Before a Browse AI trigger or action can work, the integration must be connected via an API key.

From the Automations Builder (recommended)

1. Open the Automations module and select (or create) an automation.

2. Add a Browse AI trigger or action: search for Browse AI in the Apps tab.

3. Select the desired Browse AI trigger or action.

4. On the step card, click "Connect your account."

5. In the window, paste your Browse AI API key, then click Save. The panel will then display "Connected."

To generate the key: log in to Browse AI, open your dashboard, go to Settings, then API Keys, click Create new API key, name it (for example, Automation Integration), and copy the value.

From Settings (alternative method)

1. Go to Settings, then Integrations.

2. Locate Browse AI and click Connect.

3. Paste your Browse AI API key and click Save.

Managing Multiple Browse AI Accounts

If you connect multiple Browse AI accounts, each trigger and action panel displays a “Connected Account” selector to target a specific account at each step. When only one account is connected, it is selected automatically.

[📸 SCREENSHOT: The "Connected Account" selector on a Browse AI step card in the automation builder]

The available trigger

This version offers a single trigger, whose event type is configured via filters.

The four available actions

Example: Configure the trigger (New task completed)

This workflow sets up an automation that runs every time a specific Browse AI robot completes a task, and uses the captured data as the source for all subsequent actions.

Step 1: Add the trigger

• Open the automation and click Add a trigger.

• Switch to the Apps tab and search for Browse AI.

• Select “New Task Completed” from the list of triggers.

Step 2: Configure the trigger

• Connected account: Choose the Browse AI account to monitor. It will be selected automatically if there is only one account.

• Trigger Name: Give it a descriptive name, such as “New competitor page scrape.”

• Filters, Robot: Choose the Browse AI robot to monitor. The list is populated from the connected account.

• Filters, Selection Operator (optional): Refine the selection based on a status, the value of a captured field, or a run entry. This is useful when the same bot is running on multiple entries and you’re only interested in a subset of them.

• Add filters: Apply additional conditions if the automation should only run in certain cases.

Step 3: Test the trigger

• Click Find New Records in the test panel.

• The system retrieves the most recent completed tasks that match the criteria.

• From the mapping reference selection menu, choose one of the returned records. This locks the captured data schema for the following steps.

• Click Save Trigger.

Example: Configure an action (Run a task)

This workflow launches a Browse AI robot on a URL submitted by a contact, then continues the automation once the data is available.

Step 1: Add the action

• In the automation, click Add to insert a new step.

• Open the Apps tab and select Browse AI.

• Choose "Run a Task" from the list of actions.

Step 2: Connect (first time only)

• If the integration isn’t connected yet, click Connect Your Account on the action card.

• Paste your Browse AI API key and click Save.

Step 3: Configure the action

• Action Name: A descriptive label, such as “Scrape the URL submitted by the customer.”

• Bot: Choose the Browse AI bot to run. The list is populated from the connected account.

• Input parameters: Each bot defines its own input format (a URL, a search query, a list of fields, etc.). Pass automation variables into the corresponding fields, for example, {{form.submitted_url}}.

Step 4: Proceed with the result

The Run a Task action returns a task ID. Two approaches:

• Wait using a trigger: Complete this run by storing the task ID on the contact. A separate automation, listening for “New Task Completed,” takes over when the scraping finishes.

• Wait using a delay: If the scraping finishes quickly, add a “Wait” step (for example, 30 seconds), followed by a “Retrieve a Task” action using the returned ID to retrieve the data within the same workflow.

Test Before Publishing

Test the trigger

• In the trigger panel, click Find New Records.

• If no completed tasks appear, run the robot manually in Browse AI to generate a sample, then run the search again.

• Select the returned task as the mapping reference. Verify that the captured fields (the columns your robot extracts) appear in the payload preview.

Test the action

• In the action panel, click Test Action.

• The action runs on the live Browse AI account using sample values.

• Open Browse AI and verify that the task appears in the robot’s execution history.

• If an error occurs, double-check the robot ID and the mapping of the input parameters, then run the test again. Once you’re satisfied, click Save Action and run a full test of the automation before publishing it.

Real-World Use Cases

Competitive Monitoring with Instant Alerts

Objective: Detect changes on a competitor’s page as soon as they occur.

• Trigger: New task completed (filtered by the bot that scrapes the competitor’s page).

• Action: Conditional branch based on a captured field (price, title, status).

• Action: Send an internal notification (Slack, SMS, email) if the field has changed.

Example: A bot scrapes a competitor’s pricing page according to the Browse AI schedule. Each time a task completes, the automation runs. If the price has changed since the last run, the marketing team receives a Slack message with the old and new values.

URL submitted by a contact, on-demand scraping

Objective: Launch a scraping job on a URL that the contact has just provided, then continue the automation with the captured data.

• Trigger: form submission (with a URL field).

• Action: Run a task (using the submitted URL as the bot’s input).

• Split into two automations: complete the first one by storing the task ID on the contact.

• Second automation, trigger: “New task completed”—retrieve the contact using the task ID and update its captured data.

Bulk lead enrichment from a CSV

Objective: Process a list of URLs or company names at scale and write the enriched data to the contacts.

• Trigger: Import a file (CSV with N rows).

• Action: Run a task in batch mode (one task per line, passing the values from the line as input to the robot).

• Wait for or trigger on the end of the batch.

• Action: Retrieve a batch (to obtain the aggregated results).

• Action: For each result, update the corresponding contact in the CRM.

Frequently Asked Questions