TesterArmy – AI Agents That Test Your Web and Mobile Apps

TL;DR

TL;DR: TesterArmy lets you write end-to-end tests by describing flows in plain English. AI agents handle the browser automation, catch bugs before users do, and ship screenshots, recordings, and bug reports via Slack, Discord, or straight into your pull request.

Source and Accuracy Notes

⚠️ This section is MANDATORY. All links must be verified from actual source, not guessed.

• Project page: tester.army
• Source repository: No public GitHub repo (service platform)
• License: Proprietary (YC-backed startup)
• HN launch thread: news.ycombinator.com/item?id=44725306 — Launch HN: TesterArmy (YC P26), June 18, 2026
• Source last checked: 2026-06-29

What Is TesterArmy?

TesterArmy is an agentic QA platform built by a three-person team (Oskar, Szymon, and Piotr) and backed by Y Combinator’s P26 batch. The core idea: instead of writing Playwright selectors or maintaining test scripts, you describe what a user journey should do in plain English and the agent executes it.

The platform targets web and mobile apps. You paste a staging or production URL (or upload a mobile binary), write a test in natural language, and the agent handles the rest — navigating pages, filling forms, logging in via OAuth or OTP, and evaluating results.

Key differentiator: this is a service, not a framework. There is no test code in your repo. No SDK to install. No infrastructure to maintain. The same browser primitives as Playwright underneath, wrapped in a managed agent layer.

Setup Workflow

Step 1: Create a Project

Sign up at tester.army and create a project by providing either:

• A staging or production URL for web apps
• A mobile binary (iOS or Android) for native apps

No SDK, no test scripts, no infrastructure setup required.

Step 2: Write Tests in Plain English

Describe the journey you want tested. For example:

Search results should return the correct filtered items and pagination controls must work as expected.

The agent handles clicking, typing, and verifying — just like a human tester would.

Step 3: Choose How to Run

Trigger tests from multiple entry points:

• GitHub App — connect once, get PR checks automatically on every deployment
• Production monitoring — schedule recurring runs to catch regressions in live traffic
• Webhook — fire from any CI pipeline (GitHub Actions, GitLab CI, or any tool that can hit a webhook endpoint)
• CLI — run tests on demand from the command line

Step 4: Get Reports

On every run, you receive:

• Screenshots and recordings of the session
• Actionable bug reports
• Pass/fail status in the dashboard, CLI, or directly in your pull request

Alerts route to Slack or Discord when a run fails.

Deeper Analysis

What makes it different from Playwright MCP?

Playwright gives an agent raw browser control — you still write the test logic. Playwright MCP has the same primitive problem. TesterArmy is a managed service: you describe the outcome, not the implementation. The agent maintains the test and ships evidence on every release.

Authentication handling

The agent logs in with stored credentials, handles OAuth flows, and receives one-time passwords via dedicated per-agent inboxes for OTP-protected routes. Credentials are encrypted at rest with AES-256-GCM.

Mobile support

TesterArmy tests iOS and Android builds alongside web and mobile-web experiences. Upload your binary, point at a URL, and the same natural-language test framework covers all surfaces.

Pricing

Three tiers as of June 2026:

| Plan | Price | Test runs | Projects | |------|-------|-----------|----------| | Hobby | $99/mo | 250 runs | 2 | | Startup | $299/mo | 1,000 runs | 5 | | Enterprise | Custom | Custom | Custom |

All plans include unlimited members. The Startup plan uses “the best frontier models” for test execution. A free trial gives new teams 3 test runs before upgrading.

Practical Evaluation Checklist

• No SDK to install — URL or binary is all you need
• Natural language test authoring — no selectors to maintain
• GitHub App for automatic PR checks on deploy
• CLI for on-demand runs
• Mobile app testing (iOS and Android)
• Authentication support including OAuth and OTP
• Slack and Discord alerting on failures
• Webhook trigger for any CI pipeline
• Screenshots and recordings on every run
• Free tier available (3 test runs)

Security Notes

Credentials used in tests are encrypted at rest using AES-256-GCM. The platform does not currently offer a public bug bounty or security whitepaper — enterprise customers can book a call to discuss custom security requirements.

FAQ

Q: How does TesterArmy actually test my website? A: It launches a real browser, navigates your site like a human, and uses AI to evaluate whether the expected behavior occurred. You describe what to test in plain language; the agent handles clicking, typing, and verifying.

Q: Is there a free trial? A: Yes. Every new account starts with 3 free test runs so you can evaluate TesterArmy on real projects before upgrading.

Q: How is this different from Playwright, Cypress, or Playwright MCP? A: Playwright and Cypress are developer frameworks — you write and maintain tests, debug flaky selectors, and manage infrastructure. Playwright MCP gives an agent raw browser primitives. TesterArmy is a managed service — describe the journey in English, and the platform runs it, maintains it, and delivers evidence on every release. No test code in your repo.

Q: Can I run tests in CI? A: Yes. TesterArmy integrates with GitHub Actions, GitLab CI, and any pipeline that can fire a webhook. Check statuses and report links appear directly on your pull requests.

Q: Does it handle authentication and OTP? A: Yes. The agent logs in with stored credentials, handles OAuth, and receives one-time passwords via dedicated per-agent inboxes. Credentials are encrypted with AES-256-GCM at rest.

Q: Does it support mobile apps? A: Yes. TesterArmy tests iOS and Android builds alongside web and mobile-web experiences.

Conclusion

TesterArmy tackles a real pain point: AI coding tools have made shipping code dramatically faster, but testing is still a bottleneck. Instead of maintaining fragile selectors or writing test scripts, you describe what a user should experience and the agent verifies it across web and mobile.

The service layer is the key insight — no SDK, no infrastructure, no test code in your repo. For teams that want AI-powered QA without the maintenance overhead, this is worth evaluating. The free tier gives you enough runs to test it on a real project before committing a paid plan.