Glossary of Testmode terms
The terms used across Testmode and its documentation, from action, API key and credit to finding, journey, the map, Self-driving, session replay and Viewer.
Short definitions of the terms you meet in Testmode and in these docs. Where a term is also a label in the app, it is written the way the app shows it.
| Term | Definition |
|---|---|
| Action | One operation a test runs, or its final Expected Outcome check. Each costs one credit. See How credits are counted. |
| Activity | Self-driving’s record of everything it did, with Undo for its changes to the suite. See See what Self-driving did. |
| API key | A secret starting with tmk_ that lets CI pipelines and AI agents use one project through the API. See API keys. |
| Assistant | The chat beside every page that writes test cases, plans, schedules and variables, explains your suite and speaks for Self-driving. See Ask the assistant. |
| Changes data | A test case setting: it submits, buys, deletes or sends something real, so it is left out of runs on production unless the environment allows it. See Pause testing and protect data. |
| Credential | A stored test login: username, password and optional login URL. A test with one starts signed in. See Credentials and automatic login. |
| Credit | What everything in Testmode is paid for in: one per action a test runs, 5 per assistant message, 10 per page explored, 50 per investigation, 100 per Self-driving review. See Plans and billing. |
| Default environment | The environment marked with a star: preselected in Run tests, and used by API runs that name none. |
| Draft | A run of new steps Self-driving wrote for a test, which leaves the test as it is. See How Self-driving investigates failed runs. |
| Environment | A site your tests run against, such as staging or production, defined by its base URL. See Environments. |
| Expected Outcome | What must be true on the page after a test’s last step. Testmode checks it by reading the page. |
| Exploration | A visit of one environment by the exploratory agent: it reads pages, tries things on them and ends in suggestions for your suite. See Start an exploration. |
| Finding | A fact testing found in your app, such as a failure at one step, with its triage and state. See Track what testing found. |
| Goal | What matters most to your team, in your own words, under What matters most. Every exploration, review and investigation reads it. |
| Inbox | Where Self-driving hands back what needs a decision: questions, verdicts on failed runs, exploration suggestions. See Work through the Inbox. |
| Investigation | Self-driving’s look at a failed run: it re-runs each failure and files a verdict. See How Self-driving investigates failed runs. |
| Journey | A route users take through your app, such as “Check out”, ranked Critical, High, Normal or Low, with the test cases that cover it. |
| Live browser view | The browser of a running test or exploration, shown live to editors. |
| Map | What Testmode knows about your app: its screens, the links between them, its journeys, a product summary and your team’s notes. See The map of your app. |
| MCP | Model Context Protocol, the standard AI apps use to call tools. Testmode’s MCP server is at https://. See Connect an AI agent over MCP. |
| Note | Something your team taught Self-driving about the app, kept under What it knows. Answers to its questions become notes too. |
| Operation | One thing a test does in the browser: an action on one element, a wait for something to appear, a check, opening or reloading a page, pressing a key, or a fixed wait. Testmode works out a step’s operations once and replays them. |
| Organization | A workspace with members, their roles, a plan and projects. |
| Pause testing | A project switch that stops schedules, CI runs and Self-driving at once. |
| Plan (billing) | What the organization is on: Free, Pro, Startup or Enterprise, each with its credits a month. See Plans and billing. |
| Project | Everything needed to test one application: environments, credentials, variables, test cases, plans, runs, findings, Self-driving’s map and API keys. |
| PROD | The badge on an environment marked as production. |
| Review | Self-driving thinking about the project as a whole: it ranks journeys, plans the suite, schedules it and asks what only you know. It ends with an update. |
| Role | A member’s level of access in an organization: Owner, Admin, Editor or Viewer. See Roles and permissions. |
| Run | One execution of a set of test cases on one environment. It holds one test per test case. |
| Schedule | A rule that runs a test plan automatically on one environment. See Schedule test plans. |
| Screen | One kind of page on the map: /orders/1042 and /orders/1043 are one screen, /orders/:id. |
| Self-driving | The agent that tests a project on its own: it explores, keeps the map, plans and fixes the suite and investigates failed runs. See Turn on Self-driving. |
| Session replay | The recording of a test’s or an exploration’s browser, played back on its page for 31 days. |
| Start path | The page a test opens first, relative to the environment’s URL, such as /contact. |
| Step | One instruction in a test case, in plain English. |
| Suggestion | A change an exploration proposes: a new or changed test case, a switch-off, a plan or a variable. See Review and add suggestions. |
| Test | One test case executed once, inside a run, with its own verdict, replay and logs. |
| Test case | A named list of steps and an Expected Outcome, with optional settings. See Write a test case. |
| Test plan | A named group of test cases that run together and can be scheduled. See Test plans. |
| Trial | 7 days of Pro with 2,500 credits, card required, charged nothing until it ends. |
| Triage | What a finding turned out to be: bug, intended change, test problem, flaky or environment. |
| Update | Self-driving’s account of a review: what it did and every change it made to the suite. Shown under Activity. |
| Variable | A named value test cases share, written as {{Name}}. See Variables. |