Introduction

Reelspect is an AI-powered QA automation and test management platform for online casino games and casino websites, built for operators, game providers and aggregators. It lets you automate tests with AI agents that run them in desktop and mobile browsers, on simulators and on real physical devices. This manual explains how to write tests, run them and read the results in the Reelspect web app.

How it works

You describe a test in plain language, for example “open the game, set the bet to the minimum, spin once and check that the balance goes down by the bet”. An AI testing agent then carries out the test the same way a manual tester would.

The agent works from the screen only. On every turn it looks at a screenshot, decides what to do next and acts: it clicks or taps, types, scrolls, presses keys or waits. It does not read the page’s HTML (DOM), does not use selectors and has no hooks into the game engine. Because of this, it can test canvas-based slot games, where buttons and reels are pixels rather than page elements, as well as ordinary casino websites.

When the agent finishes, Reelspect records a verdict (see Verdicts), a written explanation with expected and actual behaviour, the steps to reproduce, and a recording you can replay turn by turn.

The core flow

1. Sandbox: your private space. Draft a test case, launch it straight away and refine the instructions until the agent does what you want. See Sandbox.
2. Library: the project’s shared test cases. Promote a test case from the Sandbox when it is ready. See Library.
3. Runs: group Library test cases into runs inside test plans, choose a device, and launch them once, in several rounds, or on a schedule. See Test plans and Runs.
4. Results: check verdicts, override them when needed, replay sessions and report issues to Jira. See Results and Session viewer.

Before writing your first tests, read Writing test cases: it explains what text the agent actually receives, what it cannot do, and how to write instructions that work. Terms are defined in the Glossary; numeric limits are collected in Limits.