Assertwise MCP
QA tools your AI assistant can call
Model Context Protocol (MCP) servers so assistants in Cursor, Claude and similar clients can create cases, record runs and run accessibility checks, with EU hosting defaults.
Two servers first
Author and run from the assistant
Create suites and cases, start runs, record results, pull a short evidence summary — against your TestHub workspace when early access is live.
URL checks with fix direction
WCAG-oriented scan of a URL with ranked findings. Pairs with our Accessibility sprint when you need humans on the critical journeys.
No public install URL, no MCP registry listing, and no promise of production readiness yet. Design is in progress; build follows TestHub MVP hardening.
EU QA and eng teams already in AI editors
Teams that want the assistant to touch real test assets — with scoped auth and a clear “what not to send” policy — instead of pasting secrets into chat.
- Cursor / Claude Desktop users on SaaS or fintech products
- Teams planning TestHub early access
- Need a human-led setup? See MCP Pilot
Want early MCP access?
Tell us your editor, team size, and whether TestHub or a11y tools matter more. We’ll reply with next steps — no install link yet.