Sentinel is the open-source, self-hosted agentic QA platform. Point it at your app, and autonomous AI agents will crawl, click, and break everything before your users do. No scripts. No cloud exfiltration.
Playwright and Cypress require armies of engineers to maintain brittle selectors. Cloud AI tools like Mabl and QA Wolf demand you pipe your application's internal data through their servers.
We're a cybersecurity firm. We refuse both options. Sentinel runs entirely inside your own Docker infrastructure. Your application data, credentials, and user flows never leave your network.
Give Sentinel a URL and an LLM endpoint (OpenAI, Ollama, vLLM). One command.
sentinel run https://myapp.com
Autonomous agents map your application, fill forms, and hunt for edge cases using accessibility snapshots and LLM reasoning.
Get deterministic findings with screenshots, console logs, and accessibility violations. Fail your CI pipeline automatically.
| Community | Pro | Managed | |
|---|---|---|---|
| Price | Free forever | $399 one-time | $149/month |
| License | AGPL-3.0 | Commercial | SaaS |
| Agents | Single (sequential) | Multi-agent parallel | Multi-agent parallel |
| CI/CD integration | — | ✓ CLI + GitHub Action | ✓ CLI + GitHub Action |
| Custom assertions | — | ✓ Natural language | ✓ Natural language |
| Visual regression | Basic | Baseline management | Baseline management |
| Hosting | Docker self-hosted | Docker self-hosted | We host it for you |
| Support | GitHub Discussions | Priority Discord | SLA guarantee |
| Get Started | Buy License | Subscribe |