Automate scheduled website performance checks across devices, log results, and notify teams when thresholds are breached.
The AI agent runs scheduled tests of each website on mobile and desktop using Google PageSpeed Insights and collects Core Web Vitals (LCP, FID, CLS) and a performance score. It automatically categorizes results, logs detailed data to the Google Sheets report, and updates the main data sheet with the last checked date and status. When a site falls below configured thresholds, it generates actionable recommendations and sends alerts through Email, Telegram, Discord, or WhatsApp to the right stakeholders.
Core actions the AI agent performs to maintain site health.
Retrieve URLs and thresholds from Google Sheets.
Filter out recently checked sites and batch up to 10 per run.
Run PageSpeed Insights for mobile and desktop.
Gather LCP, FID, CLS and a performance score; categorize status.
Update the report sheet with results and update last checked date on the data sheet.
Notify teams via Email, Telegram, Discord, and WhatsApp with severity-based messages and recommendations.
This section contrasts common manual monitoring pain points with concrete outcomes achieved by the AI agent.
A simple 3-step flow that non-technical users can follow.
Retrieve URLs, device types, and thresholds from Google Sheets, then filter out items checked recently and batch up to 10 sites per run.
Run PageSpeed Insights for mobile and desktop, collect LCP, FID, CLS, and overall score, then categorize each result as excellent, needs improvement, or poor.
Write results to the report sheet, update the data sheet with last processed date, and notify stakeholders via multi-channel alerts with tailored content.
One realistic scenario.
At 02:00 UTC during a nightly run, the AI agent processes three sites. Site A returns an excellent score of 92 with mobile LCP 1.8s, FID 12ms, CLS 0.10. Site B reports needs improvement with a score of 65, LCP 2.9s, FID 35ms, CLS 0.20. Site C returns a poor score of 48 with LCP 4.2s, FID 60ms, CLS 0.30. The agent logs all results to the report sheet, updates last checked dates on the data sheet, and sends severity-based alerts via Telegram to on-call engineers for Site B and Site C, including concrete recommendations.
Roles that gain clear value from automated monitoring.
Keeps site performance within targets without manual testing
Ensures fast-loading pages to support campaigns and SEO
Monitors Core Web Vitals to protect rankings
Automates monitoring as part of reliability tooling
Gains actionable site health data for decisions
Uses historical data for dashboards and reporting
Tools connected to the AI agent to perform tests, log data, and notify teams.
Fetch mobile and desktop performance metrics (LCP, FID, CLS) and overall score for each URL.
Read website lists and thresholds; write results to the report sheet; update last checked dates on the data sheet.
Send alert emails with detailed performance data and recommended actions.
Deliver alert messages to channels with severity-based summaries and links to remediation steps.
Post notification summaries to Discord channels for on-call teams.
Push alerts to mobile groups or individuals with critical results and suggested fixes.
Concrete scenarios where automation delivers reliable outcomes.
Practical answers to common setup and run-time questions.
The AI agent runs on a configurable schedule and only tests sites that haven't been checked recently, per the threshold. Tests are batched to avoid API limits, typically processing up to 10 sites per run. You can set the interval (e.g., hourly, daily) to match your monitoring needs. Results are written to a dedicated report sheet for traceability.
Yes. Thresholds are configurable and mapped to categories like excellent, needs improvement, and poor. You can adjust the score boundaries and CWV values to reflect your site’s expected performance. The agent automatically applies these thresholds during analysis and flags breaches accordingly. Changes apply to all future runs.
The AI agent tests both mobile and desktop configurations as configured in the Google Sheets settings. It collects the same metrics for both, enabling side-by-side comparisons. Results are stored per device to support targeted optimizations. This ensures you don’t miss device-specific performance issues.
Alerts are sent via Email, Telegram, Discord, and WhatsApp based on the severity of the result. Messages include the performance score, CWV metrics, and actionable recommendations. Notifications are prioritized so critical issues are surfaced first and non-critical items are grouped for review. You can tailor recipients and channels per site.
Yes. The agent processes up to 10 websites per batch to respect API rate limits. You can increase or decrease batch size if needed, but larger batches may extend run times. Large portfolios remain manageable thanks to queueing and pacing logic. Each run updates the report sheet with per-site results.
Results are written to a dedicated report sheet and the main data sheet is updated with the last processed date and status. This creates a clear audit trail and supports trend analysis over time. Historical data enable correlation with spikes in traffic or changes in content. Data retention can be configured according to your Google Sheets setup.
Yes. The AI agent uses OAuth2 credentials to access Google Sheets and the PageSpeed Insights API securely. Access is limited to authorized accounts, and keys can be rotated. Sensitive configuration is stored in a controlled environment, and you can revoke access at any time. Regular audits of permissions help protect data integrity.
Automate scheduled website performance checks across devices, log results, and notify teams when thresholds are breached.
