SmokeOps
Turns a plain-language description of a critical user flow into an approved, schedulable browser check — no Playwright scripting required.
🔗 Visit SmokeOps
Description
Checking that a website's critical paths still work — login, add to cart, checkout — usually means either writing test scripts by hand or paying for a hosted monitoring service. SmokeOps offers a middle path: you describe the flow in plain English ("log in, then add an item to the cart"), an AI agent watches the live page and proposes the exact clicks and checks needed, you review and approve them once, and from then on the check runs entirely on its own — no AI, no internet call to a model — on whatever schedule you set.
Under the hood, SmokeOps converts the approved steps into a standalone Playwright script saved in your own repository, so checks are version-controlled like any other code rather than living inside a third-party dashboard. It integrates with GitHub Actions for cron-based scheduled runs, captures screenshots and traces automatically when a check fails, and lets you export the generated checks as plain Playwright files usable outside the tool entirely. It's local-first, open source (Apache 2.0), and free.
💬 Our review
The short version: if you want synthetic monitoring without paying a recurring SaaS fee or hand-writing Playwright, SmokeOps's approve-once-then-run-forever model is a smart trade, but it's a brand-new project (created late July 2026, one GitHub star) with no community or support behind it yet.
Compared to hosted options like Checkly or Better Stack, SmokeOps costs nothing beyond the compute you already have (a GitHub Actions runner), and because checks are saved as plain Playwright files in your repo, you're never locked into a vendor's dashboard. The catch is that you're on your own for infrastructure, uptime alerting UI, and history dashboards — things a paid monitoring SaaS bundles in. It's also unproven: at one star and a few weeks old, there's no track record of it working reliably across the messy edge cases (dynamic content, auth flows, flaky selectors) that make browser testing hard in practice. Best fit: technical teams comfortable with GitHub Actions who want free, repo-native synthetic checks and don't need a polished monitoring dashboard. Weaker fit: non-technical teams, or anyone who wants uptime alerting and history out of the box without assembling it themselves.
💰 Pricing
📊 Global score
🤖 AI-enriched data
Licence Apache 2.0, aucun coût logiciel. Le seul coût réel est l'infrastructure d'exécution (ex : minutes GitHub Actions) que l'utilisateur fournit lui-même.
Pros
100 % gratuit et open source (Apache 2.0)
Aucune dépendance à un LLM une fois les checks approuvés
Checks stockés dans le repo, versionnés comme du code
Capture automatique de captures d'écran et traces en cas d'échec
Cons
Projet très jeune (créé fin juillet 2026, 1 étoile GitHub)
Aucun support ni communauté établie
Pas de dashboard d'alerting/historique intégré — à assembler soi-même