Perforce Autonomous Testing now lets release owners define one natural-language test workflow and execute the required functional and performance tests across Perfecto and BlazeMeter, reducing duplicated test design and handoffs.
One workflow across functional and performance testing
Perfecto 26.7 introduces Perforce Autonomous Testing for organizations that use both Perfecto and BlazeMeter. Teams define the application, scenarios, environments, and success criteria in natural language; the capability generates and executes the necessary tests through the two platforms.
For release leadership, the material change is operational: test design, execution intent, and outcomes can be managed through a single workflow rather than separate functional and performance testing efforts. This can reduce inconsistent coverage definitions and make release-readiness decisions easier to defend.
Key considerations:
- Both a Perfecto license and a BlazeMeter license are required.
- Teams should define ownership for scenarios, environments, thresholds, and approval criteria.
- Existing quality gates should specify how functional and performance evidence is reviewed together.
- Pilot the workflow against representative applications before changing release policy.
MCP Server support expands AI Scriptless operations
Perfecto MCP Server now supports AI Scriptless tests through natural-language prompts. Users can create, inspect, and maintain tests; view test structure; edit commands and control flow; perform test-level operations; and access snapshots.
This gives engineering teams another controlled entry point for test maintenance. It does not remove the need for review. Prompt-driven changes can affect test intent, coverage, and the reliability of release evidence, so organizations should apply code-review discipline and clear ownership to AI-generated updates.
Practical governance actions include:
- Restrict MCP Server access according to team roles and environment sensitivity.
- Require peer review for material test-flow or validation changes.
- Record the rationale for changes to critical release tests.
- Monitor test stability after AI-assisted maintenance is introduced.
Desktop application testing joins AI Scriptless
AI Scriptless now supports desktop applications through the Desktop App platform type. After connecting a desktop computer to Perfecto, teams can execute AI-driven desktop tests, monitor their execution, and analyze results through the existing reporting model.
Reports provide execution status and duration, step-level logs, failure details, screenshots, and platform filtering. They also capture desktop-specific information such as device ID, IP address, and machine name. That level of evidence can improve incident analysis and audit trails, especially where desktop applications support regulated or internal business processes.
Desktop connectivity should be treated as an architecture and security decision. Teams need defined device ownership, network access rules, machine lifecycle controls, and retention policies for screenshots and execution records.
Reusable AI context improves consistency
Release 26.7 adds global AI knowledge through an ai.txt file in the repository. Its content is automatically added as global instructions to AI prompts, helping AI actions interpret complex or non-standard application layouts, including dashboards and graphical views.
For large test portfolios, this reduces repeated prompt guidance and can make AI-assisted behavior more consistent across contributors and pipelines. The file should be managed as a governed engineering artifact, not informal documentation.
Important constraint:
ai.txtis limited to 2,000 characters.- If it exceeds 2,000 characters, Perfecto ignores the file.
- Teams should version the file, assign an owner, and test changes against high-value workflows.
Perfecto also restores AI Scriptless commands that do not require a device under test for desktop web and virtual-device testing. This reverses a command-availability restriction introduced in 26.5, allowing affected existing tests to be edited and run as expected.
Framework and documentation updates reduce avoidable failures
Quantum 3.2.4 adds configurable failure behavior for AI User-Action and AI Validation steps, more step variations, improved AI Visual Comparison behavior, fixes for AI User-Action return data and duplicate AI report entries, plus security updates.
Teams using Quantum should validate these changes in a controlled branch before broadly updating shared pipeline templates. Configurable failure handling can alter quality-gate outcomes, so test owners and release managers should agree which conditions are blocking versus informational.
Documentation updates also clarify two areas that frequently create operational friction:
- Appium users receive updated XPath 2.0 guidance, including scenarios where Appium limits expression support and available workarounds.
- Site-to-site IPsec VPN guidance explains the existing private-network connection option and how it compares with Perfecto Connect.
These updates matter for platform architects because locator compatibility and private connectivity choices directly affect test reliability, access boundaries, and supportability.
Updated device and browser coverage requires release planning
Perfecto 26.7 expands coverage across real and virtual platforms:
- Real devices: iOS 27 Beta 2, iOS 26.6 Beta 3, iOS 26.5.2, and Android 17.
- Virtual devices: iOS 27 Beta 1 with limited capacity, and Android 17.
- Browsers on Windows and Mac: Edge 150 Stable, Chrome 150 Stable, Chrome 151 Beta, and Firefox 152 Stable.
There are implementation limits to plan for. On real devices running iOS 27 Beta 2, setting geolocation does not function properly. For virtual devices running iOS 26.4 or later, Intel-only x86_64 applications cannot run; an ARM64 build is required. Organizations seeking iOS 27.x devices in their cloud should contact Perfecto Support.
Beta operating-system coverage is useful for compatibility discovery, but it should not be confused with general-availability release certification. Maintain separate test plans, reporting labels, and decision criteria for beta-platform findings.
Recommended adoption path
A measured rollout will produce better evidence than enabling every capability at once.
- Confirm Perfecto and BlazeMeter licensing and nominate workflow owners.
- Pilot autonomous testing on a small set of business-critical web or mobile journeys.
- Define review controls for MCP Server and AI Scriptless changes.
- Add a concise, version-controlled
ai.txtfile for applications with complex UI context. - Validate desktop connectivity and reporting against security and audit requirements.
- Test Quantum 3.2.4 and new platform versions in non-production pipelines.
- Update release criteria for known iOS beta, ARM64, and capacity constraints.
How Merito helps
Merito helps enterprises assess where Perfecto 26.7 fits within existing QA, DevOps, and security controls. We can define adoption roadmaps, govern AI-assisted test changes, design Perfecto and BlazeMeter workflows, validate connectivity patterns, and establish reporting that gives release leaders defensible quality evidence.
Merito is an authorized Perforce reseller and services partner. Our Perforce Perfecto team can scope licensing, sizing, and rollout for 26.7, and our enterprise upgrade services help you plan and validate the upgrade with minimal disruption to release schedules.