Browseai

@browseai get the last 10 tasks from my product-price-tracker robot and show me which ones failed

Retrieved 10 recent tasks from robot 'product-price-tracker'. Of these, 2 tasks failed: Task #847392 (failed at step 'extract price' due to page structure change) and Task #847385 (timeout after 30s waiting for page load). The remaining 8 tasks completed successfully, capturing price data from target e-commerce pages. Each successful task includes timestamps, captured screenshots, and extracted data fields.

This example shows how the MCP surfaces task execution history and failure diagnostics. The Get Robot Tasks tool returns paginated results — if your robot runs frequently, you may need to filter by date range or status. Task details include captured data and screenshots, which can help debug extraction issues when page layouts change.

@browseai create a monitor for my competitor-homepage robot that runs every 6 hours and alerts me when content changes

Created monitor 'competitor-homepage-watch' for robot 'competitor-homepage'. Schedule: every 6 hours. The monitor will automatically execute the robot and compare results against the previous run. You'll receive notifications via your configured webhooks when Browse.ai detects changes in the extracted data fields. Monitor ID: mon_8x3k9p2 — you can update the schedule or delete this monitor later if needed.

Monitors automate robot execution on a schedule and detect data changes. This example demonstrates the Create Monitor tool's ability to set up recurring checks. Note that monitor frequency affects your task quota — a 6-hour interval means 4 runs per day. You'll need to configure webhooks separately (via Create Webhook) to receive change alerts outside the Browse.ai dashboard.

@browseai I need to scrape 500 product URLs from my spreadsheet — set up a bulk run and estimate how long it'll take

Prepared bulk task run for 500 URLs using your 'product-scraper' robot. Based on the robot's average execution time (12 seconds per task) and Browse.ai's concurrency limits, estimated completion: 25-35 minutes. The bulk run will process tasks in parallel batches. You can monitor progress via the Get Robot Tasks endpoint — tasks will show status 'pending', 'running', or 'successful'. If any tasks fail due to page changes or timeouts, you'll see error details in each task's result.

The Bulk Run Tasks tool handles up to 50,000 tasks per request, making it efficient for large-scale data collection. This example shows how the MCP can help estimate run duration based on robot performance. Be aware: bulk runs consume your monthly task quota quickly, and failed tasks still count toward usage. Always test your robot on a few URLs before launching a large bulk run to catch extraction issues early.

When Browseai wins for weekly competitor price checks

A 6-person e-commerce team needs to track 200 competitor SKUs every Monday morning before pricing calls. Browseai is the right call here because you train one robot to scrape a competitor's product page, then bulk-run 200 tasks with different URLs in a single API call. The Create Monitor tool sets the weekly schedule, and the webhook fires when data changes—your Slack channel gets a ping if a competitor drops their price. This works until you hit the 50,000-task bulk limit or need sub-hourly checks (Browseai's minimum interval is hourly). If your competitors block scrapers aggressively, you'll burn through proxy budgets fast. For teams running 200-2,000 SKUs weekly with stable target sites, Browseai beats writing your own Puppeteer scripts. Book a demo if you need the robot-training UI instead of raw HTML parsing.

When Browseai handles sales prospecting at scale

A 3-person sales team closes 40 deals a quarter and needs LinkedIn profile data for 500 inbound leads monthly. Browseai is the move if those leads come from a consistent source (say, a directory or review site) where the page structure doesn't change weekly. You train a robot to extract company size, location, and contact info, then bulk-run tasks with a CSV of profile URLs. The Get Task Details tool lets you poll for completion and dump results into your CRM. This breaks down if the target site requires login (Browseai's auth handling is brittle) or if you're scraping 10 different site layouts (you'd need 10 robots). For teams doing 500-5,000 leads a month from 1-3 stable sources, Browseai is faster than hiring a VA. Try the free tier with 50 tasks before committing.

When Browseai automates candidate pipeline alerts

A 2-person recruiting shop tracks 15 niche job boards for senior engineering roles in fintech. Browseai is the right tool because you set up one robot per board, configure monitors to check every 6 hours, and get webhook pings when new postings match your keywords. The Get Robots List and Get Robot Tasks tools let you audit what's running and pull historical data for quarterly reports. This falls apart if the boards change their HTML weekly (you'll retrain robots constantly) or if you need real-time alerts (hourly is the fastest Browseai goes). For teams monitoring 5-20 boards with stable layouts and needing 6-24 hour latency, Browseai beats manual checks or building a scraper fleet. Start with 3 boards on the free plan to test reliability before scaling.