Automations from MCP and the API, domain-scoped basic auth, and knowledge file uploads

Highlight
Featured this release
Set up automations from MCP, the API and chat
Agents can list, create, update, delete and trigger automations through MCP, the public API and the in-app chat.
MCP clients get five new tools: list_automations, create_automation, update_automation, delete_automation and trigger_automation. The same actions are available over the public API at /v1/automations. The in-app chat assistant can create, edit, pause, run and delete automations, and shows each schedule in plain language with its last few runs.

Pausing an automation clears its next run, and resuming schedules the next run from that moment. One-time schedules use a pinned cron expression like 30 9 14 10 *.

More on this: MCP, API testing and smoke testing.
# Direct link
Highlight
Featured this release
Domain-scoped basic auth for environments
Protect staging and preview sites with HTTP Basic Auth rules that QA.tech applies only to the domains you choose.
Add one or more rules – domains, username, password – to an environment, and QA.tech sends basic auth only to matching hosts, with *.example.com wildcards supported. Set rules in the environment form or as a default for Preview Environments; saved passwords are write-only and show as "Stored password".

Rules apply to AI testing in the browser, API tests and the crawler, and carry through preview creation and promotion. CI integrations and the public API accept a basicAuth field on environment overrides, and the chat agent can create and update environments with basic auth. The existing "Use for Basic Auth" checkbox and user:pass@ URLs keep working.

More on this: multi-environment testing, authentication testing and integrations.
# Direct link
Update 03 of 05
New docs: the regression review agent
A new docs page explains how QA.tech reviews test plan results and what each verdict means.
The "Regression review agent" page under Regression Testing in the docs covers what gets a review (test plans started from Run Tests, schedules, CI or the API), how the agent investigates failures and reruns tests to check its hypotheses, and how verdicts are grouped: Issues, Setup – not a code bug, Flaky – passed on retry, and Tests to update. It also shows where the review appears and how to start one yourself with /regression in chat.

More on this: docs and regression testing.
# Direct link
Update 04 of 05
PR reviews list follow-up todos
Review comments show each open todo for the pull request as a tip under the verdict.
The review agent now files follow-ups it finds during a pull request review and shows them as a tip under the verdict, marked "after merge". In the todo detail they carry an "Awaiting merge" badge. When the PR merges, the post-merge review keeps or drops each one; if the PR closes without merging, pending todos are dropped.

More on this: PR testing and shift-left testing.
# Direct link
Update 05 of 05
Upload files to your knowledge library
Add Markdown, text, CSV, JSON, YAML and PDF files to the knowledge library, or ask the chat agent to add one.
Under Add Knowledge there's a new "Upload Files" option next to Web Link and Text Content. Drop in several files at once, and each becomes its own library item with an editable title. Supported types are .md, .txt, .csv, .json, .yaml/.yml and .pdf – text files up to 5 MB, PDFs up to 100 MB.

You can also attach a file in chat and ask the agent to add it to the library. It starts indexing right away, and the agent can then read added PDFs when it answers questions.

More on this: product overview, file upload testing and AI test generation.
# Direct link

Set up automations from MCP, the API and chat

Domain-scoped basic auth for environments

New docs: the regression review agent

PR reviews list follow-up todos

Upload files to your knowledge library

Your team moves fast. Can your testing keep up?