BrowserAct

@BrowserAct list all my running tasks from the last 24 hours and show me which ones are taking longest

Found 4 active tasks from the past day. Two are still running: 'competitor-pricing-scan' (started 3h ago, currently navigating page 47 of 120) and 'form-submission-test' (started 45m ago, waiting for CAPTCHA). One completed successfully ('daily-screenshot-batch', 2h runtime), and one failed ('login-flow-test', authentication timeout after 8 minutes). The pricing scan is your longest-running task and may benefit from pagination adjustments.

This example shows BrowserAct's task-monitoring capability. The List Tasks tool supports filtering by status and time range, but the response shape depends on how many tasks exist in your workspace. Tasks in 'running' state consume execution credits until completion or cancellation.

@BrowserAact cancel the competitor-pricing-scan task immediately, it's stuck in a loop

Task 'competitor-pricing-scan' (ID: task_abc123) has been canceled. The browser session terminated at page 47, and no further credits will be charged. Partial results captured before cancellation remain accessible via Get Task for the next 30 days. Note that canceled tasks cannot be resumed—you'll need to start a new run if you want to complete the workflow.

Demonstrates the Cancel Task action, which is permanent and immediate. Useful when a workflow hits an unexpected state or you need to stop credit consumption. The MCP requires the task ID (obtained from List Tasks or a previous run reference), and cancellation works even if the task is already failed or completed.

@BrowserAct search the workflow templates for anything related to e-commerce product scraping, then explain which one would work best for a site with lazy-loaded images

Found 3 relevant templates: 'E-commerce Product Catalog Scraper' (handles pagination and basic lazy-load via scroll triggers), 'Dynamic Content Extractor' (waits for network idle and DOM mutations, better for React/Vue apps), and 'Visual Product Grid Scraper' (screenshot-based, no DOM parsing). For lazy-loaded images, the Dynamic Content Extractor is your best fit—it monitors network requests and waits for image elements to populate before extraction. The Product Catalog template may miss images that load after initial scroll. Both require you to specify CSS selectors for the product containers.

This synthesis example pairs BrowserAct's template search with reasoning about technical requirements. The List Workflow Templates tool returns metadata and input parameters, but the AI must interpret which template matches the user's constraint. Template selection depends on the target site's architecture, which the MCP cannot inspect without a live run.

When BrowserAct beats manual price checks at 10+ SKUs

A 6-person e-commerce team needs daily competitor pricing on 40 SKUs across three marketplaces. BrowserAct's workflow templates handle the scraping and proxy rotation, returning structured data the team can pipe into Sheets or a dashboard. The API_KEY auth means anyone on the team can trigger a run from Switchy without touching credentials. This works until you hit rate limits or need sub-hourly updates—at that scale, you're better off with a dedicated scraping service. The 11 tools give you task lifecycle control (cancel, status, list) and workflow discovery, so you can build a repeatable cadence without writing scraper code. If your pricing window is daily and your SKU count is under 200, BrowserAct keeps the workflow in one place and the data fresh.

BrowserAct for sales teams running outbound at 50-100 leads per week

A 3-person sales team pulls LinkedIn and company-site data for 60 leads each week before cold outreach. BrowserAct's workflow templates automate the profile visits and data extraction, returning contact details and recent activity in a structured format. The Get Task and List Tasks tools let the team monitor progress in Switchy and retry failures without leaving the workspace. This setup breaks down if you need real-time enrichment during a call or if your lead volume spikes past 500 per week—then you want a dedicated enrichment API with sub-second latency. The proxy region support (List Proxy Regions) helps avoid blocks when scraping at volume. If your outbound motion is weekly batches and you're under 200 leads, BrowserAct keeps enrichment simple and the data in one shared context.

When BrowserAct handles quarterly audit trails for small compliance teams

A 4-person compliance team at a fintech needs timestamped screenshots of 30 public-facing pages every quarter for SOC 2 evidence. BrowserAct's workflow templates capture the pages, store the files, and return metadata the team can attach to audit reports. The Get Task tool surfaces live-view details and file outputs directly in Switchy, so the team doesn't juggle storage buckets. This works for quarterly or monthly cadences; if you need daily snapshots or version-diffing, you'll want a dedicated monitoring tool with alerting. The Cancel Task and List Tasks tools give you recovery options when a run hits a timeout or a page layout changes mid-capture. If your audit frequency is quarterly and your page count is under 100, BrowserAct keeps the evidence collection repeatable without custom infrastructure.