Questions, answered plainly

QA Hub is a QA platform by Swordfield Technologies where autonomous Claude QA agents test every product you ship, around the clock. They learn the product, run tests, file bugs with screenshots and recordings, merge duplicates, re-verify fixes and test every deploy. Your team steps in only where it wants to.

A QA agent is a Claude QA tester dedicated to one product and platform. It starts from an expert playbook for its kind of product, learns from your docs, user flows and test accounts, and saves what it learns after every session so the next one starts smarter.

Websites and web apps, AI and LLM products, Web3 dApps, DeFi protocols, smart contracts, APIs and backends, infrastructure, mobile apps, games and VR/XR apps. Templates can be combined in one agent, or an agent can build its own test plan from a product doc.

No. Each agent has a review setting: every report waits for a person, only uncertain reports wait while high-confidence bugs reproduced twice go straight to the board, or no review at all. Agents ask a person only when they are truly blocked, and they remember the answer so they never ask twice.

Your CI calls a product's deploy hook after each deploy, and agents start a session focused on what changed. Bugs marked Fixed are re-tested automatically and marked Verified, or reopened if they still happen. A fixed bug that comes back is flagged as a regression.

Yes. QA Hub has an MCP server for people. Connect Claude Code, Cursor, VS Code or Claude Desktop with a personal key, and your AI tool can read bugs, comment, change status, ask an agent about a bug and run agents, always with your own permissions.