# MCP for AI coding tools, smarter onboarding, and chat-attached tests > 12 updates: MCP – Connect AI coding tools to QA.tech with one command, Agent selector is now top-level; crawl trigger removed, Android testing now… Source: https://qa.tech/changelog/2026-05-26 --- 12 updates shipped on May 26, 2026. In this release 1. 01[MCP – Connect AI coding tools to QA.tech with one command](#2026-05-26-mcp-configure-cli)Highlight 2. 02[Agent selector is now top-level; crawl trigger removed](#2026-05-26-agent-selector-top-level) 3. 03[Android testing now defaults to Pixel 9 Pro / Android 15](#2026-05-26-android-defaults-pixel-9-pro) 4. 04[Captcha detection and solving improved](#2026-05-26-captcha-detection-solving-improved) 5. 05[GitLab integration settings now explain the @mention trigger](#2026-05-26-gitlab-mention-trigger-docs) 6. 06[Live status cards for background agent operations](#2026-05-26-live-status-cards-background-agents) 7. 07[Onboarding chat now researches first, then proposes](#2026-05-26-onboarding-chat-researches-first) 8. 08[Open any test case directly in chat](#2026-05-26-open-test-case-in-chat) 9. 09[PR review summary redesigned with linked test names](#2026-05-26-pr-review-summary-redesigned) 10. 10[Project knowledge lookup rebuilt](#2026-05-26-project-knowledge-lookup-rebuilt) 11. 11[Slack run notifications now show trigger context and results](#2026-05-26-slack-run-notifications-context) 12. 12[Small fixes](#2026-05-26-small-fixes) 1. Highlight Featured this release ## MCP – Connect AI coding tools to QA.tech with one command @qadottech/cli is on npm, and the API has 16 agent-callable tools – wire them up with npx @qadottech/cli mcp configure. Run `npx @qadottech/cli mcp configure -c cursor` to write QA.tech's [MCP server](https://qa.tech/product/mcp) into Cursor's config automatically, or `-c claude-desktop` for Claude Desktop. `--print` always outputs the snippet if you'd rather paste manually. Re-running is idempotent and won't remove other entries from your `mcpServers`. The same snippets are available under Settings → Integrations → API in a new "MCP clients" section. The API now exposes 16 agent-callable tools via `POST /v1/mcp`, including `start_run`, `create_chat`, `list_application_environments`, and `create_remote_tunnel`. [\# Direct link](#2026-05-26-mcp-configure-cli) 2. Update 02 of 12 ## Agent selector is now top-level; crawl trigger removed The "Agent" dropdown is now a top-level setting in the test case editor. The "Triggers crawl" toggle has been removed. You no longer need to expand "Advanced" to switch which agent runs a test case. The crawl trigger had no effect on normal runs and has been removed. [\# Direct link](#2026-05-26-agent-selector-top-level) 3. Update 03 of 12 ## Android testing now defaults to Pixel 9 Pro / Android 15 New test cases default to Google Pixel 9 Pro with Android 15 instead of Pixel 8 / Android 14. Camera and barcode scanning are available on Android 15+; mid-test rotation works on both. Existing presets are unaffected. iOS defaults remain iPhone 15 Pro / iOS 17. [\# Direct link](#2026-05-26-android-defaults-pixel-9-pro) 4. Update 04 of 12 ## Captcha detection and solving improved The agent now handles captchas more reliably during test execution. Detection and solving logic has been rewritten with updated methods. Standard image/audio captchas and reCAPTCHA variants are now handled more reliably. Cloudflare Turnstile is still not supported. [\# Direct link](#2026-05-26-captcha-detection-solving-improved) 5. Update 05 of 12 ## GitLab integration settings now explain the @mention trigger The GitLab settings page links to docs and explains you can trigger a review by commenting @qa.tech with a preview URL. Comment `@qa.tech ` on any MR to trigger a QA review, without needing native GitLab environments. The settings page now surfaces this alongside the auto-run toggle and environment mapping. [\# Direct link](#2026-05-26-gitlab-mention-trigger-docs) 6. Update 06 of 12 ## Live status cards for background agent operations Test generation, crawls, runs, and PR review posting now show a live status card inline in the chat. Background agent operations – test generation, crawls, runs, PR review posting – now show a live status card inline in the chat. Previously those operations appeared to silently hang while running. The card updates in real time and can be dismissed when done. Cancellation works from the same card across all operation types. [\# Direct link](#2026-05-26-live-status-cards-background-agents) 7. Update 07 of 12 ## Onboarding chat now researches first, then proposes The setup chat reads your site before asking anything, generates the obvious tests, then offers concrete choices. The agent acknowledges with one line, silently reads the crawl and application details, then generates the top flows (login, primary paths, adjacent flows) without waiting for your answers. From there it offers two to four specific directions to go deeper – named flows, not open questions. Each choice kicks off another research wave. The final step also now redirects immediately instead of waiting for the full response to complete. [\# Direct link](#2026-05-26-onboarding-chat-researches-first) 8. Update 08 of 12 ## Open any test case directly in chat Right-clicking a test or pressing CMD+K now shows "Open Test in Chat", opening a new chat with that test attached. The chat opens with a reference card showing the test's name, owner avatar, and application badge. The agent immediately has the test in context. The action is hidden when multiple tests are selected. [\# Direct link](#2026-05-26-open-test-case-in-chat) 9. Update 09 of 12 ## PR review summary redesigned with linked test names The bot's summary is now a short narrative with test names linking to their results, instead of a table. The "Other issues worth notice" table is gone. The new format has a bold verdict, one paragraph per failure with a linked test name, then a paragraph naming what passed. The structure is the same for approved and not-approved reviews. [\# Direct link](#2026-05-26-pr-review-summary-redesigned) 10. Update 10 of 12 ## Project knowledge lookup rebuilt Documentation search now answers questions rather than returning raw matches, and fixes timeouts on large knowledge bases. The old search timed out on large docs sets. It's replaced with an "ask" interface that synthesizes answers directly. Project and documentation knowledge are now queried through a single tool. [\# Direct link](#2026-05-26-project-knowledge-lookup-rebuilt) 11. Update 11 of 12 ## Slack run notifications now show trigger context and results Run notifications include who triggered the run, a result summary, and action buttons – not just pass/fail. Notifications show the test plan name, a trigger line ("Triggered from PR #N", "Triggered via API", "Scheduled run", etc.), and a count like "5 passed · 2 failed". Chat and GitHub PR triggers add "Open conversation" and "Open PR" buttons. [\# Direct link](#2026-05-26-slack-run-notifications-context) 12. Update 12 of 12 ## Small fixes Tracer dark mode, autoplaying assessment videos, integration logo proportions, and legacy pricing cleanup. - **Tracer dark mode:** API call view (method badges, headers, media toggle) is now readable on dark backgrounds. - **Assessment videos autoplay:** Videos in the assessment report view now start playing automatically. - **Integration logos:** Non-square logos (e.g., Netlify) on the connections page now render at correct proportions. - **Legacy pricing:** Old plan types no longer appear in billing settings. [\# Direct link](#2026-05-26-small-fixes) ## Your team moves fast. Can your testing keep up? QA.tech agents test your product autonomously, so moving fast never means shipping broken. See how it works in a 30-minute demo. [Get a demo](https://qa.tech/demo) [QA.tech](https://qa.tech/) SOC 2 Type 2 compliant ![Sensiba SOC 2 Type 2 compliant](https://qa.tech/assets/soc2-type2-badge-DPb1j9VX.png) ### Product - [Platform overview](https://qa.tech/product) - [AI QA Testing](https://qa.tech/ai-qa-testing) - [Web Testing](https://qa.tech/product/web-testing) - [Mobile Testing](https://qa.tech/product/mobile-testing) - [PR Testing](https://qa.tech/product/pr-testing) - [MCP](https://qa.tech/product/mcp) - [Use cases](https://qa.tech/use-cases) - [Changelog](https://qa.tech/changelog) ### Compare - [QA.tech vs Cypress](https://qa.tech/compare/qa-tech-vs-cypress) - [QA.tech vs Playwright](https://qa.tech/compare/qa-tech-vs-playwright) - [QA.tech vs Mabl](https://qa.tech/compare/qa-tech-vs-mabl) - [QA.tech vs QA Wolf](https://qa.tech/compare/qa-tech-vs-qa-wolf) - [AI vs Human QA](https://qa.tech/saas) ### Resources - [Docs](https://docs.qa.tech/) - [Blog](https://qa.tech/blog) - [Events](https://qa.tech/events) - [Pricing](https://qa.tech/pricing) - [Case Studies](https://qa.tech/case-studies) ### Company - [About](https://qa.tech/about-us) - [Partnerships](https://qa.tech/partnerships) - [Press](https://qa.tech/press) - [Contact Us](https://qa.tech/contact) - [Privacy Policy](https://qa.tech/privacy-policy) - [Terms of Service](https://trust.qa.tech/resources?s=mb9w4yjb03i3t8bjsq4jx3&name=main-agreement.pdf) - [Security & Compliance](https://trust.qa.tech/) ### Social - [](https://www.linkedin.com/company/qa-tech-ai/) - [](https://x.com/getqatech) - [](https://www.instagram.com/getqatech/) © 2025 QA.tech. All rights reserved