Checkly

Monitoring used to mean "can the public internet reach this?" That's not enough. The services that matter to your business also need to be monitored: internal admin panels, partner portals, finance tools, the legacy app some intern built ten years ago... they never see the public internet. In our May 5 session, Daniel demo-ed Checkly Private Location. "With Private Locations in Checkly, you can monitor things that are literally not possible to monitor any other way." Recording here: https://hubs.ly/Q04fLGLM0

María de Antón had nine failing tests on screen. Chromium, Firefox, mobile Chrome. Instead of opening the trace, she said: "I'm gonna ask Rocky." One click. Rocky AI reads the Playwright trace, the network logs, the screenshots, and the OpenTelemetry backend spans already captured for that run. Then it tells her exactly what broke. This is what AI-native debugging looks like when the AI lives inside your debugger. Watch the full session: https://hubs.ly/Q04hNFdr0

O11yCon is on! We'll be at our table running demos all day and would genuinely love to connect. Stop by for a live demo of what AI-native reliability actually looks like: Claude Code writing Playwright tests, `npx checkly deploy` putting monitors live across the globe, and Rocky AI doing the 2 AM thinking so on-call doesn't have to. Come tell us what's been breaking lately. We'll show you what we've been building. Our team will be there and ready to chat. 📍Convene 100 Stockton, San Francisco 🔗 https://hubs.ly/Q04g4mq60 See you there 👋

O11yCon is on! We'll be at our table running demos all day and would genuinely love to connect. Stop by for a live demo of what AI-native reliability actually looks like: Claude Code writing Playwright tests, `npx checkly deploy` putting monitors live across the globe, and Rocky AI doing the 2 AM thinking so on-call doesn't have to. Come tell us what's been breaking lately. We'll show you what we've been building. Our team will be there and ready to chat. 📍Convene 100 Stockton, San Francisco 🔗 https://hubs.ly/Q04g4mq60 See you there 👋

Tomorrow we’re going live with our webinar on The AI-Native Playwright Reporter. If your team is already running npx playwright test in CI/CD, GitHub Actions, or locally, this session is for you. We’ll show how the Checkly Playwright Reporter helps teams move beyond guesswork with: - Rocky AI analysis on real Playwright failures - error grouping and trends over time - richer debugging context with traces, network views, console logs, and performance graphs - a path from test results into production monitoring Join María de Antón and Pırıl Kavlak tomorrow for a live demo and see how Playwright teams can better understand what broke, what keeps breaking, and how to debug faster. Register here: https://hubs.ly/Q04ftxQT0

When something breaks in production at 2 AM, how fast can your team find it? Instead of just waking up to an alert and working to triage it, you can wake up to the cause and possible fixes. Checkly is the active reliability layer your agents can operate themselves. They write the Playwright tests, deploy the monitors globally, get root cause + impact analysis from Rocky AI, and open the PR with the fix; all through the CLI. No dashboards required. We'll be at O11ycon 2026 with a live demo of the loop: → Agent ships a feature → Agent writes the tests → `npx checkly deploy` puts monitors live worldwide → When something breaks, Rocky AI returns the root cause, user impact, and a suggested fix — straight to the agent If you're building toward AI-native reliability, come find us. We would love to connect in person! 🗓 May 20–21 · 📍 Convene 100 Stockton, San Francisco 🔗 https://hubs.ly/Q04g4nxR0

I’m really excited to be joining María de Antón for our upcoming Checkly webinar on May 21! 🎭 We’ll be talking about the AI-Native #Playwright Reporter and showing how Playwright teams can stop guessing and debug failures, flakiness, and trends with more context. We’ll cover: -> Rocky AI analysis (our agent) on real Playwright failures -> grouped errors and recurring issues over time -> richer debugging context in Checkly -> and how this workflow can extend into production monitoring If you’re already running npx playwright test in CI/CD, GitHub Actions, or locally, this is for you. 🎉 Register here: https://lnkd.in/dq7E-ids

Your Playwright test failed. Now what? For most teams, that question still means opening CI logs, digging through traces, checking screenshots, rerunning the test, and trying to piece together what actually happened. On May 21, we’re hosting a live webinar on The AI-Native Playwright Reporter to show a better flow. We’ll walk through how the Checkly #Playwright Reporter helps teams: - use Rocky AI analysis on real Playwright failures - group recurring errors and track flaky behavior over time - debug faster with traces, network views, console logs & performance context - connect test results to production monitoring If your team runs npx playwright test in CI/CD, GitHub Actions, or locally, you’ll want to see this live. Register here: https://hubs.ly/Q04ftll60

Every company has a few apps like this. A legacy internal tool, half the team depends on it, nobody wants to touch it. Standalone EC2, no SSL, no API, never going to be public. Most monitoring stops at the perimeter, so these apps are the biggest blind spot in your reliability stack. When they break, the first signal is usually a Slack message from someone furious it's down. In our May 5 session, Paulus shows a check running from a Checkly Private Location. Same reliability layer you have on your public services. Full demo: https://hubs.ly/Q04fM9-H0

Letting an agent open incidents and deploy monitoring sounds risky. So we built the guardrails into the CLI, not the agent: `incidents create`, `deploy`, and `destroy` return exit code 2 with a confirmation envelope when an agent calls them. The Checkly skill teaches the agent to surface the diff and wait for human approval before retrying. In the April 28 session, Stefan let Claude push a fix and open a status page incident. Both calls hit the confirmation gate. Both required a yes before going through. YOLO behavior is structurally blocked, not vibes-blocked. Watch Stefan walk through it: https://hubs.ly/Q04g2NJT0