Changelog

Codex, Cursor, and Ollama now support the same tool access, effort levels, and plan-mode workflow as Claude Code in the TestRelic CLI, so you can drive a TestRelic-aware AI agent from whichever engine your team already uses.

Ask AI conversations are now budgeted by actual usage instead of a fixed session cap, and your organization's quota is re-checked throughout a long-running session rather than only at the start — so longer investigations no longer cut off early.

The API session workspace's Cookies view now shows the request and response cookies captured for each API call, with sensitive values automatically masked before they ever leave your machine.

You can now turn on two-factor authentication or register a passkey for your own account from Settings, and organization admins can require it for every member.

Ask AI can now begin a conversation from a specific test case or screenshot instead of a blank prompt, and produces shareable artifacts you can attach to a ticket or share with your team.

The MCP server's tool list was trimmed by over 20% so assistants start faster, all supported AI assistants now get the same guidance out of the box, and small version mismatches between the server and your assistant no longer break the connection.

Very large test runs no longer risk being rejected outright — oversized reports are now trimmed in stages so you always get a usable report instead of a failed upload.

Very large test runs no longer risk being rejected outright — oversized reports are now trimmed in stages so you always get a usable report instead of a failed upload.