Documentation › Glossary
Glossary
If a word in the docs is unfamiliar, it is probably defined here. Cross-references link to the fuller explanation.
Core objects
| Term | Definition |
|---|---|
| Project | An isolated workspace containing its own cases, suites, schedules, files, parameters and users. All of a customer's work lives inside one or more projects. Every user is also a member of the shared community examples project. |
| Case | A single automated test. A case has a name, a type/language (MXML, AI, Java, Python, JavaScript, Robot Framework), a platform (Web, Android, iOS), a script and a set of parameters. See the User Guide. |
| Suite | A named group of cases (and/or other suites) that run together, in sequence or in parallel, and produce a consolidated Allure report. Suites can contain suites. |
| Page object | A reusable model of a single screen/page, referenced by cases so that locators live in one place. Supports Java, MXML and UTAM. |
| Schedule | A rule that runs a case or suite automatically on a recurring timing (minutely, hourly, daily, weekly, monthly), e.g. a nightly regression run. |
| Run | One execution of a case or suite. Every run is recorded in the Runs list with a result, state, duration, logs, console output and (optionally) a video. |
| Template | A starting point for a new case/page object — a JSON object bundling script code, language, framework and description. Stored in the file system's templates folder. |
Case types & languages
| Term | Definition |
|---|---|
| MXML | MaxTAF XML — MaxTAF's own XML-tag language, specialised for testing IBM Maximo. It transpiles to Java before execution and handles Maximo's quirks (iframes, dynamic fields) with dedicated tags such as <ui:click> and <ui:assertValue>. See the MXML Reference. |
| AI test case | A case whose "script" is plain-English instructions. When run, MaxTAF drives an AI agent that controls a real browser to carry out the steps, and can record the interactions into a reusable MXML case. See Writing AI Test Cases. |
| UTAM | UI Test Automation Model — an open, YAML/JSON-based page-object format (originating from Salesforce). In MaxTAF a UTAM page object compiles into two page objects (Java + JavaScript). |
| Robot Framework | A keyword-driven test framework layered on top of the Python structure; cases are .robot files. |
ApiService / mxService | The MaxTAF library object available inside a case (one per language) that lets the case talk to MaxTAF — read parameters (mxService.getParam), read/write project parameters, take screenshots, add report lines and share values between runs. |
Execution & infrastructure
| Term | Definition |
|---|---|
| Engine | The MaxTAF component that compiles and runs cases for a project. Its health is shown by the Engine dot in the bottom status bar; it can be restarted from Project Settings ▸ Main. |
| Node | A browser worker that actually executes a test in a real browser. Nodes are drawn from a pool. |
| Common pool | The shared pool of browser nodes (${mx.system.common_pool}). Runs queue when the pool is busy; free plans have no parallelism. |
| Private pool | Dedicated, on-demand nodes for an account (${mx.system.private_pool}) — for parallelism and isolation. |
| Playwright grid | The execution tier used by AI test cases: a hub distributes AI runs to nodes, each of which runs an AI agent that drives a browser through a Playwright MCP. See Writing AI Test Cases. |
| MCP | Model Context Protocol — the tool interface an AI agent uses to control the browser (navigate, click, type, evaluate) during an AI test case. |
| Selenium / WebDriver | The browser-automation protocol MaxTAF uses to drive Chrome/Firefox/Edge for scripted (non-AI) cases. Configured via mx.selenium.* parameters. |
Parameters
MaxTAF is heavily parameter-driven. A parameter is a name/value pair read by a case. Standard parameters are reserved and prefixed mx.; custom parameters are user-named variables. See the full Parameter reference.
| Term | Definition |
|---|---|
| Parameter layer | The scope at which a parameter is set: test, suite, user, project or system. Layers combine according to mx.params.source and mx.params.method. |
| Project parameter | A value shared across every case in a project (e.g. an environment URL). Managed on the Dev ▸ Parameters page; read with mxService.getProjectParam("NAME") or referenced as ${mx.project.NAME}. |
| Configuration parameter | A project parameter flagged for auto-injection into runs. Required for mx.recorder.* and similar keys to reach the engine. |
| Data-driven testing | Running a case once per row of a spreadsheet (CSV/Excel), mapping columns to parameters via mx.datadriven.*. |
| Intersection / union | The two ways parameter layers merge (mx.params.method): union combines all layers; intersection keeps only parameters present in a chosen target layer. |
Reporting & tooling
| Term | Definition |
|---|---|
| Allure report | A rich graphical test report generated automatically for suite (and case) runs, viewable in-app or as a pop-out. |
| Live video / take control | A live feed of the browser during a run; the mouse icon lets you take manual control mid-run. |
| The recorder | The MaxTAF Chrome extension that captures your interactions with a web app and generates an MXML case. See the Recorder Guide. |
| AI Assistant | The in-app AI chat helper (Learn ▸ chat icon) that helps you write and debug test code — distinct from an AI test case. See the User Guide. |
| Bin | The recycle bin for deleted cases/suites — deleted items can be restored from here. |
| Bridge | MaxTAF Bridge — an optional component that lets cases execute against a browser on your own machine's local Selenium server. |