Release Radar Ep06 | Five Vendors Gave AI a Role

AI agents are showing up in enterprise delivery tools the same way a new team member shows up for work. They get a role, their work gets flagged, and what they ask for gets checked. In Release Radar Episode 06, Chris Carpenter walks through 8 releases published between October 7 and October 22, 2026 from OpenText, TestRail, Perforce, Black Duck, Semgrep and Sonatype, on screen and one page at a time.

The observation this month is that 5 vendors started treating AI like a user role. Every agent is one more account, and someone should (needs to?) approve what it can touch.

Starting with Merito UFT Phoenix

The episode features something Merito built. Merito UFT Phoenix converts OpenText Functional Testing tests, the tool most teams still call UFT, from VBScript to Python, in place in an OpenText ALM project or from a folder of tests on disk.

VBScript stays fully supported and switching to Python is optional, but it is now available in OpenText Functional Testing. For teams that choose Python, Phoenix starts with a read only pass that makes zero changes and returns a GO or NO-GO verdict with every blocker named. Converted tests keep their test IDs, links and Test Lab history. This two-page customer brief covers the migration, the Functional Testing upgrade and the OpenText licensing Merito can run alongside your team, and the installer is available through request access.

Where AI got a role this month

OpenText SDP 26.4 makes its MCP server officially available, so an AI client can read, edit and create items and add comments. Teams that used the MCP server before 26.4 need a space admin to grant OAuth consent and assign MCP permissions to the right roles. The release also brings plain language defect search that matches on meaning, which makes duplicate defects easier to catch.

Perforce Perfecto 26.10 combines AI User Action and AI Validation into one AI Command and controls access per person through the AI User role. Appium 3.6.0 becomes the Perfecto default on January 3, 2027, the same date Appium 1 extensions need to move to a supported version.

TestRail 10.9.1032 adds is_ai_generated and is_ai_automated to its API, so a report can separate tests your team wrote from tests a tool wrote. It also adds single-call reruns and automatic deactivation for inactive users.

Semgrep October 2026 brings Guardian into Codex with the fixed guardian-default ruleset, and makes Agentic Workflows scheduling and run triggers an org admin task.

Sonatype Nexus Repository 97 introduces Pro+, which checks components requested by developers, builds and AI agents before they reach a build. Base telemetry is now required, with repository restrictions after a grace period.

Perforce BlazeMeter 4.1 adds an MCP plugin for JMeter, so the MCP servers your agents depend on can be load tested like any other production service.

Also covered

Black Duck Detect 12.1.0 adds OpenJDK 25 support and drops Java 8. Sonatype IQ Server 208.1, released October 22, stops trusting root certificates affected by the eMudhra certificate authority compromise.

Every release covered in this episode is written up in full under product release updates.

Somebody on your team has to decide what an agent is allowed to touch, and who signs off on it. If that is the problem you are working on, that is the work we do. Book a conversation HERE.

Merito is a Value-Added Partner delivering AI-powered, secure, efficient and lasting enterprise solutions in AppSec, Quality, and DevSecOps.

New episode on the first and third Tuesday.