QA Agent That Writes Test Cases, Runs Them, and Self-Heals Automation
AI support across multiple testing phases
UiPath applies AI to requirement evaluation, test-case generation, test-data generation, test automation, result analysis, reporting, and maintenance. Autopilot can generate manual, low-code, coded, and API-oriented test cases from natural-language instructions. Agent Builder allows teams to build their own specialised testing agents.
Broad enterprise application coverage
UiPath states that Test Cloud supports testing for more than 190 enterprise applications and technologies, including SAP, Salesforce, Epic, and Oracle. It also supports web, desktop, mobile, API, Citrix, and other enterprise application scenarios.
Low-code and coded authoring
Tests can be created visually using UiPath activities or programmatically using C# or VB. This makes the platform accessible to automation testers while still allowing experienced developers to implement more complex logic, reusable libraries, API tests, and data-driven testing frameworks.
Integrated test management
Test Manager provides requirements traceability, test-case management, test sets, execution results, defect tracking, reporting, and connections to external ALM systems. Tests can be linked to requirements and defects, giving organisations a central view of coverage and test status.
Strong support for enterprise UI automation
UiPath’s established UI automation capabilities can be reused for application testing. The platform supports browser automation, Windows applications, Citrix and remote applications, SAP interfaces, mobile devices, and applications for which conventional DOM-based automation is difficult.
Custom AI agents
Agent Builder allows testing teams to create purpose-specific agents without developing an entire AI infrastructure. UiPath provides examples including agents that retrieve test data, consolidate failed tests into distinct bugs, summarise and distribute results, and check test cases for compliance with organisational standards.
Self-healing capabilities
The Enterprise plan includes Healing Agent, which can identify broken UI selectors and attempt to repair test automation after an application interface changes. This can reduce the maintenance burden associated with selector-based UI tests.
Flexible execution
Tests can run using cloud robots or customer-managed infrastructure. Serverless cloud robots can scale web and API performance tests automatically, while on-premises robots can be used when tests require desktop access, internal systems, or local infrastructure.
Enterprise governance
Test Cloud includes centralised access control, test and AI governance, data-handling controls, analytics, audit capabilities, and regional deployment options. Enterprise customers can obtain dedicated or on-premises deployment, bring their own AI models and encryption keys, and use their own credential vaults.
The product architecture is complex
Agentic Testing involves several UiPath products rather than one application. A complete implementation may include Test Cloud, Test Manager, Studio or Studio Web, Orchestrator, Robots, Autopilot, Agent Builder, Healing Agent, Integration Service, and external ALM tools. Administrators must understand how these services, users, robots, licences, folders, tenants, and consumption units relate to one another.
The licensing model can be difficult to understand
UiPath uses separate platform, user, robot, and consumption-based licences. Relevant licences include App Tester, App Test Developer, and App Test Robot, while some AI capabilities consume Platform Units, Agent Units, Heals, or ScreenPlay Runs. The total cost therefore depends on the number and type of users, execution infrastructure, AI usage, and deployment model.
Public pricing is not available
UiPath does not publish a fixed price for Test Cloud Standard or Enterprise. Both require a sales quotation, which makes early comparison with tools offering transparent per-user or per-execution prices more difficult.
Advanced desktop testing remains Windows-oriented
Studio Desktop, which is required for the most advanced coded, data-driven, and complex desktop scenarios, is installed on Windows. Studio Web works on Windows and macOS and supports cross-browser and native macOS application testing, but it does not provide every advanced feature available in Studio Desktop.
AI-generated tests still require preparation and review
For UI test generation, UiPath recommends creating Object Repository elements that represent the application’s UI controls. The quality of generated tests also depends on the completeness of requirements, acceptance criteria, workflows, mock-ups, and prompts. Generated tests should therefore be reviewed rather than treated as automatically correct.
Mobile testing requires additional infrastructure
UiPath mobile automation is based on Appium. Teams must configure local devices, emulators, an Appium server, network-accessible devices, or an external cloud device farm. UiPath orchestrates and automates mobile tests but does not itself replace a large hosted real-device service.
Some important AI capabilities are plan-dependent
Built-in AI agents are included in Test Cloud Standard, but self-healing automation, bring-your-own-model support, extended ALM integration, dedicated deployment, and several advanced governance capabilities are associated with Enterprise licensing. Feature availability also differs between commercial Test Cloud, Public Sector, and Dedicated deployments.
It may be excessive for small browser-testing projects
Teams that only need lightweight browser test generation and execution may find UiPath broader and more complicated than necessary. Its strongest value is in enterprise environments involving multiple applications, legacy systems, desktop interfaces, business processes, and centralised governance.
Autopilot is UiPath’s prompt-based generative AI capability for testing. A tester can use natural-language instructions to create, modify, refactor, or analyse tests in Test Manager and Studio.
Its principal testing capabilities include:
Requirement evaluation
Autopilot can analyse requirements for clarity and quality. It can help identify incomplete, ambiguous, inconsistent, or insufficiently testable requirements.
Manual test-case generation
Autopilot can generate manual test cases and test steps from requirements. Users can provide additional context such as process diagrams, application mock-ups, compliance documents, meeting transcripts, and acceptance criteria.
Missing scenario identification
It can suggest test scenarios that may be missing from the current test portfolio, including positive, negative, boundary, and end-to-end scenarios.
Low-code test generation
Autopilot can turn manual test steps or natural-language instructions into UiPath low-code test cases composed of automation activities.
Coded test generation
It can generate C#-based test automation using UiPath coded automation APIs. It can also assist with code generation, expressions, refactoring, and test logic.
API test generation
Autopilot can help create and modify API-oriented tests. Studio itself supports importing API definitions, including Swagger-based definitions, and building automated tests for services.
Test-data generation
Autopilot can create synthetic test data and data variations for data-driven tests. Generated data can be reviewed and refined before it is used in test executions.
Fuzzy verifications
UiPath can use AI-assisted fuzzy verification when an assertion should allow semantic or approximate matching rather than an exact deterministic match. UiPath lists fuzzy verification generation among the Studio AI capabilities available to testers.
Test-result analysis and reports
Autopilot can analyse selected execution results and generate a downloadable report. Test reports generated through this feature can be exported as DOCX files.
Autopilot Chat in Test Manager
Autopilot Chat can interact with Test Manager functions, execute supported commands, evaluate requirements, generate test cases, and invoke external automation processes or agents. It can also connect to custom MCP servers.
Agent Builder enables teams to create custom AI agents for testing processes. UiPath presents the following example agents:
Agents can be created using low-code tools. UiPath also supports coded agents that can be developed in an external IDE and connected to the UiPath platform using an SDK.
Healing Agent provides self-healing UI automation. When selectors fail because an application interface has changed, the Agent can analyse the context and attempt to locate the intended UI element using alternative information.
Healing Agent is available in commercial Test Cloud and Test Cloud Dedicated but is not currently listed as available in Test Cloud Public Sector. Its use is consumption-based through Heals and Agent Units.
ScreenPlay is UiPath’s agentic desktop automation capability. It interacts with application interfaces more adaptively than conventional selector-only automation and can navigate interfaces, interpret context, and respond to changes. UiPath’s Test Cloud documentation identifies ScreenPlay as one of the capabilities that can support adaptive, AI-assisted test design and execution.
ScreenPlay should not be confused with a deterministic testing framework. Because its behaviour can be more adaptive, teams must define appropriate validations and governance around agent-driven interactions.
UiPath provides central controls for AI usage and data handling. Enterprise plans include bring-your-own-model capabilities for supported AI products. Model and feature availability may differ by deployment region and by Test Cloud edition.
Test Manager supports manual test cases, requirements, test sets, execution records, evidence, defects, and requirements traceability. Manual tests can later be converted into low-code or coded automation.
UiPath supports functional testing of browser, desktop, remote, Citrix, SAP, and other graphical applications. Tests can use selectors, computer vision, image-based automation, semantic activities, keyboard and mouse actions, and reusable Object Repository components.
Web tests can be created through Studio Desktop or Studio Web. UiPath supports visible and headless browser execution, depending on the browser and automation method.
Windows desktop applications can be tested through Studio Desktop and Windows Robots. UiPath can automate traditional desktop software, Java applications, SAP GUI, remote desktops, virtualised applications, and applications that expose limited conventional automation interfaces.
Studio Web additionally supports native macOS application testing, although advanced desktop scenarios may still require platform-specific tools and activities.
UiPath Studio supports API test automation, including the import of API definitions and the creation of service-level tests. API tests can be combined with UI, database, and business-process validations in the same test workflow.
UiPath supports:
Mobile automation uses Appium as the underlying driver.
Test Cloud supports performance testing by reusing automated functional tests as virtual-user scenarios. Web and API loads can run through serverless cloud robots, while desktop performance testing requires on-premises robots.
This approach is particularly relevant for business-process performance testing because a virtual user can execute a realistic sequence across application UIs and APIs rather than only sending protocol-level requests.
Tests can run with multiple data variations from Excel files, JSON test-data files, UiPath Data Service or Data Fabric entities, Orchestrator test-data queues, existing project data, and AI-generated data.
UiPath provides specialised support for SAP and SAP S/4HANA testing. More broadly, it claims compatibility with more than 190 enterprise technologies, including Salesforce, Epic, and Oracle applications.
UiPath can test UiPath workflows and RPA processes in addition to external business applications. Test cases can validate individual activities, workflow components, integration points, and complete automation processes. Testing robots may run test cases and automation processes in non-production environments.
The learning curve can be described as moderate for basic test management and low-code automation, but high for complete enterprise implementation.
Test Manager provides a conventional test-management interface for requirements, manual tests, test sets, and results. Studio Web requires no local installation and is intended as the quicker entry point for browser and macOS testing. Autopilot can help users create initial test scenarios, test data, and automation workflows.
UiPath Academy provides free courses, including a course specifically designed to introduce Agentic Testing with Test Cloud to automation developers who have limited or no previous UiPath experience.
For complex projects, teams may need to learn:
Autopilot reduces some authoring effort, but it does not remove the need to understand the tested application, test architecture, assertions, execution infrastructure, and result interpretation.
The Test Cloud and Studio Web interfaces officially support:
Internet Explorer, 32-bit browsers, mobile browsers, unsupported browsers, and older browser versions are outside the supported scope.
UiPath’s browser-extension automation mode supports:
WebDriver-based visible and headless testing supports Chrome, Edge, and Firefox. Chromium Automation supports Chrome, Edge, and other Chromium-based browsers but not Firefox or Safari.
Consequently, the exact feature set is not identical across all four browsers. Some activities, headless execution modes, extensions, and interaction methods are browser-specific.
Windows
Studio Desktop is installed on Windows and provides the broadest support for coded tests, desktop automation, complex workflow logic, data-driven testing, and enterprise UI automation.
macOS
Studio Web works on macOS and can be used for cross-browser web testing and native macOS application testing. Local browser automation is supported through compatible macOS Robots and browser extensions.
Linux
The Test Cloud interface can be accessed through a supported browser, and cross-platform or serverless workflows can execute tasks that do not require Windows desktop interaction. However, conventional Windows desktop UI testing requires Windows-based execution infrastructure.
Android and iOS
UiPath supports real and virtual Android and iOS devices through Appium-based mobile automation.
UiPath can test:
Serverless robots are suitable primarily for web and API testing and do not support desktop automation. Desktop tests require an appropriate VM or on-premises Robot.
Test Manager offers native or out-of-the-box integrations for systems including:
The exact list can differ by Test Cloud, Automation Cloud, Automation Suite, and standalone version.
For more complex integrations, UiPath recommends Test Manager Connect, which is provided through Planview Hub. It supports configurable, bidirectional synchronisation between Test Manager and external ALM systems.
Compared with older native connectors, Test Manager Connect supports bidirectional synchronisation, custom issue types, custom fields, configurable field mappings, and more complete markup translation.
UiPath’s Enterprise pricing information states that Test Cloud can integrate with more than 50 ALM tools.
UiPath explicitly identifies support for:
These integrations can build, deploy, execute, and report UiPath tests as part of CI/CD pipelines. Test execution can also be controlled through UiPath command-line tools, APIs, Orchestrator, and custom pipeline scripts.
UiPath supports integration with SAP Cloud ALM and SAP Solution Manager. Automated test cases can be created, synchronised, executed, and monitored between SAP’s lifecycle tools and UiPath Test Manager.
Test Manager exposes APIs, SDKs, scopes, external-application registration, and webhooks. These can be used to synchronise requirements, defects, stories, automation packages, and execution results with custom development and testing tools.
Autopilot Chat can connect with custom MCP servers, allowing it to invoke external tools, automation processes, or AI agents from Test Manager.
UiPath offers three formal support levels.
Bronze Support is included with product licences. It provides basic technical assistance for UiPath products, robots, and agents.
Gold is a separately purchased support tier. It includes:
Platinum includes the Gold services and adds strategic advisory support, deployment and governance guidance, security assistance, regular automation-program health reviews, and an assigned Technical Account Manager.
Customers and evaluators also have access to:
The Customer Portal supports technical case creation and 24×7 access for eligible Premium Support customers.
UiPath does not publish a fixed Standard price. Customers must contact UiPath Sales for a quotation.
The Standard plan includes:
Individual users, robots, and AI or execution consumption may nevertheless require separate licences or consumption units.
Enterprise pricing is also available only by quotation.
Enterprise adds:
UiPath identifies three principal Test Cloud roles:
Depending on the features used, organisations may also consume:
For example, Healing Agent first consumes allocated Heals and may subsequently consume Agent Units, depending on tenant configuration.
Test Cloud does not currently offer self-service purchasing or self-service trial registration. Trials must be arranged through UiPath Sales.
The standard Application Testing trial runs for 60 days and includes:
The publicly advertised $25-per-month Basic plan applies to UiPath’s broader entry-level Automation Cloud offering. It should not be interpreted as the price of Test Cloud Agentic Testing. Test Cloud Standard and Enterprise both require a customised sales quotation.
The tool has been successfully submitted. We will add it after reviewing it.
Our excellent customer support team is ready to help.
