What is AegisRunner?

AegisRunner explores web and native mobile products, generates and independently reviews tests, runs eligible cases, and keeps findings tied to evidence.

AegisRunner is a managed testing platform for web and native mobile products. Instead of beginning with a blank test file, it begins with the running application.

Explore. A real browser or device follows reachable screens, controls and workflows within the scope and budget you choose.
Generate. An authoring model creates positive, negative, boundary and multi-step cases from observed product evidence.
Review. A different model family checks test depth and missing controls, then adds cases without overwriting the original generation record.
Run and explain. Eligible cases are replayed, and results retain the page, screen, action and author/reviewer provenance that produced them.

What a scan can produce

Generated functional test cases for reached pages and workflows.
Page-grouped functional failures, broken links and visual findings.
Accessibility, SEO, security-header, performance and API evidence where those checks apply.
Screenshots, browser logs and request evidence for investigation.
Explicit refused or unsettled coverage when a capability cannot be verified safely.

Coverage is bounded by the page limit, navigation depth, authentication, application structure and time budget. AegisRunner does not claim that a scan reaches every route or proves every business rule.

Testing behind login

Add login details from the project setup or scan flow. AegisRunner can use declared form-based login steps and recover from a session drop during longer scans. Projects with multiple roles can save separate login profiles.

Account sign-in, password reset and 2FA are available directly on app.aegisrunner.com; they do not require opening a project first.

Read-only and mutation-capable testing

You decide whether testing may submit forms or perform state-changing actions. Mutation-capable testing requires explicit project authorization. Without that consent, the execution fence blocks unsafe HTTP and WebSocket writes while allowing the declared login request through a request-bound one-shot permit.

A blocked or ambiguous mutation does not become a passing test. It remains refused, inconclusive or not tested.

Independent test review

The primary model authors test cases. A separate reviewer model checks depth and control coverage. Generation, review additions and runtime outcomes are recorded by generation, so teams can measure whether reviewer-added cases find failures the original author did not.

The reviewer is not a substitute for human product judgment. People still define business intent, approve destructive testing, assess risk and investigate ambiguous evidence.

Web and mobile

Web tests run through Chromium, Firefox and WebKit profiles. Native mobile testing accepts iOS and Android builds and runs on managed real-device infrastructure. Cross-surface journeys can connect a mobile action to a web verification when the created identity can be grounded safely; unsupported or ambiguous identity capture fails closed.

What you can do with generated tests

Run selected suites from the UI.
Schedule supported runs.
Trigger scans and runs from CI with project-scoped tokens.
Review page-grouped issues and their evidence.
Export generated web suites as Playwright/POM code on Pro and Business.

Current plan limits

Plan	Projects	Pages per scan	AI test coverage	Scans / runs per month
Free	1	25	1 page / scan	2 / 5
Starter	3	100	Up to 100 pages / scan	150 / 500
Pro	10	500	Up to 500 pages / scan	500 / 2,500
Business	25	Unlimited	Unlimited pages / scan	2,000 / 10,000

See Pricing for current prices and Subscription Plans for feature details.

Get started

Run a one-page public scan without an account.
Create a Free account.
Follow the Quick Start Guide.
Inspect live demo evidence.