1. Docs
  2. Help
  3. Glossary

Glossary

Every word you meet on a SmartestQA screen, from A to Z. Each entry says what it is in one or two sentences and links to the page that explains it in full.

22 min read Updated CustomersAdmins

In short

On this page you can

  • Look up any on-screen term, from Action to Zephyr Scale.
  • Tell apart words that look alike, such as test agent and autonomous agent.
  • See what DEV, TEST and PRP, dry run and pipeline run mean in practice.
  • Jump from each term to the page that explains it step by step.

How to use it

  1. Click a letter at the top, or search the page for the word you saw.
  2. Read the one or two sentences of the entry.
  3. Follow its link for the procedure, the formula or the fix.
  4. Need the big picture first? Read Core concepts, then come back.

A · B · C · D · E · F · G · H · I · J · K · L · M · N · O · P · Q · R · S · T · U · V · W · X · Z

New to SmartestQA? Core concepts explains how these terms fit together.

A#

Access token (repository)
An optional read-only token for your Git provider. With it, Deploys & Repos also shows the commits and changed files between two versions. Connection modes
Action
One control on one page, such as Click Sign In on login, with its type, locators, optional Navigates To and screenshot. Test steps point at actions, so one fix reaches every test. Actions
Action List
The Discover tab with every action of the project in a searchable table, where you edit, fix and clean up controls and locators. Action List
Active (user)
Whether a user can sign in. An inactive user keeps their history but cannot sign in, and still counts against the user limit. Deactivate a user
Active Directory
Signing in with your company’s directory user name and password. Directory groups decide each person’s role and projects. Active Directory
Admin
The role that can do everything in the company: projects, users, Project Panel, Company Settings, Integrations and Logs. Whoever signs up is the first Admin. Roles & permissions
Agent API key
The key of one test agent (host). The agent, the local MCP server and CI jobs use it to act with that user’s access. API key
AI account
Which AI provider a project uses: Smartest AI (SmartestQA’s account, the default) or Own Account (your provider and key). AI Account
AI Analyze
The button on a failed step in a report that asks the AI to classify the failure and suggest a cause. Analyze a failed step
AI Options
The Project Panel section with Enable AI, the AI account and one switch per AI capability. AI Options
AI Report Feedback
An AI review of a whole run’s results, shown as a card on the report summary. AI Report Feedback
AI step (AI Ask, AI Check, AI Do, AI Chat)
A step that tells the AI what to do in plain words: answer a question about the screen, check a statement, perform an action, or reply without the screen. AI steps
AI verdict
The AI’s label for a failed step: Real Bug, Test / Infrastructure Issue, Flaky or Unclassified, with a confidence. Verdicts
API request
One saved HTTP call (method, URL, headers, body, authorization) in the API Requests library. API Requests
API step
A step that calls an HTTP API inside a test case. It keeps its own copy of the request it was made from. API Request step
App Library
Your company’s APK, AAB, IPA and simulator builds, ready to install on Device Farm devices. It holds up to 30 apps. App Library
App Map
Every page as a record: purpose, preconditions, expected outcomes, risk, roles, screenshots and where it leads. The AI reads it. App Map
Archive
Where archived test cases, folders and requirements go until they are restored or deleted for good. Archive and restore
Assertion
A step that compares an actual value with an expected one (Equals, Contains, Greater Than…) and fails the step when they differ. Assertion step
Assessment (AI)
An AI review of a main requirement and its sub-requirements, with a Quality Score from 0 to 100 and suggestions. AI assessment
Attempt
One execution of one test case inside a run. Each data row and each automatic retry is an attempt. Report building blocks
Audit log
The record of every change in your company: who made it, from the app, MCP or an API key, and whether it went through. Admins only. Audit log
Authorization (API)
How a request signs in: None, Bearer Token, Basic Auth or Inherit from parent. Test runs apply only Bearer Token (Authorization in runs). Authorization
Auto-fix
The Action List button that fills missing Navigates To values from passing test runs. Unreachable pages
Auto-heal
When every locator of a step fails, the runner finds the element again, saves the new locator and leaves a repair for review in Maintain. How auto-heal decides
Auto Task Creation
A run setting: after a pipeline run, a task is created for each failed test case. Automatic tasks
Autonomous agent
An AI job in the Agents menu that runs on a schedule, after a deploy or on demand, and leaves a report and proposals. Not the same as a test agent. Autonomous Agents
AWS Device Farm
Amazon’s device cloud, one of the places a mobile project can run its tests. AWS Device Farm
Azure DevOps
An issue tracker integration: tasks become work items, and runs can be published to Test Plans. Azure DevOps

B#

Before and After
Pipeline phases that run before and after each Test row, for example to open a start screen or close dialogs. Phases
BrowserStack
A device cloud a mobile project can run its tests on. BrowserStack
Budget blocked
An autonomous agent run that stopped because the agent’s daily budget was used up. Agent limits

C#

Cases (badge)
How many test cases use an action, shown in the Action List. Click it to list them. Cases badge
CI/CD integration
Starts a SmartestQA pipeline from your build, waits for it, returns a JUnit report and fails the build when tests fail. CI/CD
Collection
The top level of the API Requests tree, usually one per API or service. Every import creates one. API building blocks
Company
Your SmartestQA account. It owns the projects, the users and the licence. Company settings
Concurrent runs
How many server runs your company may have at once in the shared pool. Admins see it in Network and capacity. Network and capacity
Confirmed by runs
A locator repair that worked in three more runs and is accepted without a person. Confirmed by runs
Confluence
An integration that publishes run reports and test case documentation to Confluence Cloud pages. Confluence
Connected actions
The Discover actions a Data Source key belongs to. They tell the AI which pages need which data; values are never sent. Connected actions
Connection (DB or Kafka)
A database or Kafka login stored in Data Source, used by Database and Kafka steps. The password lives here, never in a test. Connections
Connection test
A read-only call to another system with the credentials you entered, to check them before saving an integration. Integration building blocks
Connector (claude.ai)
The hosted SmartestQA MCP server that claude.ai connects to over the internet, with your SmartestQA sign-in. Connect claude.ai
Continue On Failure
A run setting. Off: after a test case fails, the rest of the run is skipped. On: the next test case runs. Test Settings
Cron (Custom period)
A five-field expression (minute, hour, day, month, weekday) for a pipeline schedule that Daily, Weekly and Monthly cannot express. Cron

D#

Daily summary
One chat message a day with yesterday’s runs and today’s open tasks, at the hour you pick. Daily summary
Data Source
The project’s store of keys, datasets and database and Kafka connections that steps read by name. Data Source
Database step
A step that runs an SQL query against your database and saves column values as variables. Database step
Dataset
A table uploaded from XLSX, XLS or CSV, or typed in. Steps read a column as {{excel:name}}_column, and a pipeline can run once per row. Datasets
Dedicated server
A SmartestQA test server that serves only your company. When it is full, runs can move to the shared pool. Server choice
Deploy
A record your CI sends when a version of a repository reaches an environment, with its commit, status and time. Versions and history
Deploy token
A secret starting with sdt_ that lets your CI record deploys for one repository. It is shown only once. Deploy token
Desktop application
An app a DESKTOP project’s test cases can open, with its path or ID. One is the default. Applications
Desktop Recorder
Captures the controls of a Windows or macOS desktop app through the agent, for DESKTOP projects. Desktop capture
Device
One phone, tablet, emulator or simulator reported by a node, shown as Available, Busy or Offline. Devices
Device Farm
The menu item for test agents and, in mobile projects, the nodes, devices and App Library of SmartTest Device Farm. Agent Pool & Device Farm
Disconnected
A run or attempt that stopped reporting: still Running more than 2 hours after it started, usually because the agent slept or went offline. The 2-hour rule
Discover Your App
The WEB scan that captures every interactive control of a page, with locators and screenshots, page by page. Discover Your App
Documents folder
A separate root in Analyze for project documents such as specs. It cannot hold requirements. Documents
Dry Run
A trial run of one test case from the Automate editor. Live log only: it writes no report and does not change the status icon. Dry run and live log

E#

Egress IP
An address from which Run on Server runs reach your app. Add it to your firewall allow list (Our egress IPs). Egress IPs
Enable AI
The main AI switch of a project in AI Options. With it off, every AI feature of the project is off. Enable AI
Encrypted key
A Data Source key whose value is stored encrypted, shown only to its creator and read only by test runs. Use it for every password and token. Encrypt a key
Environment (DEV, TEST, PRP)
Which installation of your app a recording or run opens. PRP is the production-side environment. Start page and environment
Environment-specific key
A Data Source key ending in _DEV, _TEST or _PRP, used instead of the plain key in that environment. Environment variants
Environment mapping
Which environments you expect a repository in, with an optional branch and start URL. It decides which deploy cells show Not mapped. Environment mapping
Execution (InSight)
The InSight tab that shows how runs went in a period: summary cards, failure clusters, flaky and slow test cases. Execution
Execution History
The list of a pipeline’s past runs with their status and a link to each report. Execution History
Execution mode (mobile)
Where a mobile project’s app runs: your agent’s device, SmartTest Device Farm, AWS Device Farm or BrowserStack. Mobile Config
Expand Scenarios
An AI action in Automate that turns one test case into negative and edge-case variations. Expand scenarios
Export (project)
Download a project as JSON, as a Java Selenium project or as a TypeScript Playwright project. Export

F#

Failure cluster
Failed steps whose error messages are the same once numbers and quoted values are ignored. One cluster often has one cause. Failure clusters
Farm run
One test run on one Device Farm device, listed in Recent runs with its status, duration and device metrics. Recent runs
Flaky score
How often a test case’s result flips between runs: 0 is stable, 1 flips every time. Flaky score
Folder
Groups test cases in Automate and requirements in Analyze. A folder directly under the project folder is a module. Folders

G#

Generate Steps
An AI action that writes a test case’s steps from a prompt, reusing your recorded actions. It replaces the steps in the editor. Generate from a prompt
Get
A step that reads a value from an element and saves it as a variable, with optional transforms such as To Number. Get
Global parameter
A value stored under a project-wide name. The Global parameter step sets one; values that other steps produce are stored too. Global parameter step

H#

Healed (tag)
An amber tag on a test case while at least one of its auto-heal repairs waits for review. Healed tag
Host
Another name for a test agent, used in dialogs such as New Host and in API key messages. Test Agents

I#

InSight
The home dashboard, with the Execution and Test Health tabs. InSight
Integration
The saved settings for one outside system in one project, such as Jira or Slack. Admins set them up. Integrations
Iteration
A repeat of the same test case with the next data row or execute count, shown with a Run selector in the report. Report building blocks

J#

Jira
An issue tracker integration: tasks become Jira issues, with statuses and priorities mapped both ways. Jira
JMeter
A load-test step that runs a .jmx plan, uploaded or generated by AI from the test case’s API steps. Load test steps
JSON path
How an API step saves a value from the response into a variable, for example the id of a created order. Extract values
JTL file
A JMeter results file. Report reads it to show the results of a load run. JTL files

K#

k6
A load-test step that runs a k6 .js script, uploaded or generated by AI. Load test steps
Kafka step
Publishes a message to a Kafka topic, or waits for a matching one, to test event-driven flows. Kafka steps
Key (Data Source)
A named value such as loginEmail, unique in the project. Steps read it as {{ds:loginEmail}}. Variables (keys)

L#

License Details
The dialog with your licence type, dates and usage limits such as Max User Count and Max Thread Count. License Details
Licence limit
The most your company’s licence allows, such as users, test cases or parallel threads. Counters count what you create; deleting does not give capacity back. At a limit you see “Company license limit has been reached…”. Billing
Live log
The Test Run panel that streams while a run is going. It is not kept; reports come from pipeline runs. Report record and live log
Live session
You take a phone or emulator and see and control its screen in the browser until you release it. Live device sessions
Load test scenario
A test case that holds only JMeter, k6 and API steps. Recording and browser steps are not available in it. The load-only rule
Localhost agent
Your personal test agent, named after your e-mail address. Every user has one; it cannot be renamed or deleted. Agent list
Locator
How a test finds a control: id, name, css, xpath, accessibility_id, resource-id and others. An action can keep fallback locators. Discover building blocks
Locator chip
The chip in the Action List with the main locator’s type and the number of fallbacks, coloured by type. Locator chip

M#

Mail step
Reads a one-time code or link from an e-mail, waits for a message, counts messages or sends an e-mail during a test. Mail steps
Main requirement
A feature or user flow in the Analyze tree, such as Pay with a saved card. Its sub-requirements hold the testable conditions. Create a main requirement
Maintain
The menu item with the review queues: Auto-heal, AI Steps and Network Calls. Maintenance
Manage URLs
Opens the list of start page URLs to add or edit one, with an address per environment. Start page URL
Mapping (integration)
The link between a SmartestQA item and its copy in another system, such as task 412 and Jira issue PAY-88, so syncs do not create duplicates. Integration building blocks
Maturity
A weighted average of the four Test Health pillars, from 0 to 100. Maturity
MCP
The Model Context Protocol. The SmartestQA MCP server lets Claude and other assistants read and change your project, showing each change before it is written. What MCP is
Members (task)
The people assigned to a task. A task can have several; none means unassigned. Members
Mobile Config
The Project Panel section that decides where a mobile project’s app runs and how. Mobile Config
Module
A business area such as Orders. It groups pages in the App Map and is the top-level folder in Analyze and Automate. Modules

N#

Navigates To
The page an action leads to, such as Click Sign In → dashboard. These links draw the App Map. Discover building blocks
Network Calls
The Maintain tab that turns the browser calls a Network step captured into a new API test case. Network Calls
Network step
Checks, reads or lists the calls the browser makes during a web test: status codes, response fields, durations, JS errors. Network step
Node
A machine that connects phones and emulators to the Device Farm, with its own health checks. Nodes
Not tested (deploy)
A deploy cell with no pipeline run in the same environment after the deploy. Cell states
Note
One rule or fact about the app that no scan can see. Active notes are sent to the AI with every generation for their pages. Notes

O#

Online and Offline (agent)
An agent is Online while its heartbeat arrives; it shows grey and Offline when none arrived for 90 seconds. When an agent counts as online
Own Account
Using your company’s AI provider and key for a project, instead of Smartest AI. Use your own account

P#

Page
A name that groups actions, such as checkout. Pages are created by naming them. Page names
Parallel Execution
Runs the test cases of a pipeline at the same time, up to the Thread Count. Parallel Execution and Thread Count
Period
When a pipeline starts by itself: Real-time (never), Daily, Weekly, Monthly or Custom (Cron). Period
Pillar
One of the four Test Health coverage percentages: requirement, scenario, risk and action coverage. Pillars
Pipeline
A named, saved run plan: a run list plus its settings. It runs by hand, on a schedule, from CI or from a trigger, and every run writes a report. Schedule
Pipeline Parameters
The values a pipeline’s steps type or select, changed for that pipeline only. The test case keeps its Case default. Pipeline Parameters
Playwright
One of the two browser engines of WEB projects, the default for new projects. Playwright settings
Project
One application under test, of type WEB, MOBILE or DESKTOP. Everything you build belongs to one project. Create a project
Project access
The projects a User may open. Admins see all projects; a User with none sees “No Project Access”. Give project access
Project Panel
Configuration › Project Panel: the settings of the selected project. Admins only. Project settings
Project type
WEB, MOBILE or DESKTOP. It decides how you capture the app and where its tests run. Which project type?
Proposal (autonomous agent)
A change an agent wants, saved with its exact input. Apply runs it with your permissions; Dismiss drops it. Proposals
Proposal and preview (MCP)
A change an assistant shows as a diff, with a preview id. Nothing is written until you approve and the preview is applied. Propose, then apply
Proving scenario
A test case with at least one step that checks a result, such as an assertion or an API extractor. Only these count as coverage in Test Health. InSight building blocks
PRP
The production-side environment. Tests on PRP change real data, so run only agreed read-only checks there. Start page and environment
Purpose (page)
What an App Map page is for, written by you. The Setup Roadmap counts Describe the app as done at 60% of pages with a purpose. Page information

Q#

Quality Score
The assessment’s 0–100 rating of how clear and complete a requirement’s text is. It does not say whether the requirement is tested. Quality Score
Queue (Schedule)
A list of work items; each item is one run of its processing pipeline with the item’s data filled in. Queues
Queued (server)
A server run waiting for a free slot. It never times out and starts when its turn comes. When a run waits for a server

R#

Record Actions
Captures only the controls you use, in flow order, instead of scanning whole pages. Record Actions
Repair (auto-heal)
One change the runner made during a run, shown as a card in Maintain › Auto-heal: a new locator, or an assertion accepted despite a format difference. The repair card
Report
The stored record of one pipeline run: verdict, test cases, steps, screenshots, logs and video. Report
Repository
A Git repository connected in Deploys & Repos, with a provider, a service name and an optional access token. Connect a repository
Result (chip)
The passed and total attempts of a run in the Reports list, such as 5/6. Data rows and retries each count. Result chip
Retry Count
How many more times a failed test case is run in a pipeline, from 0 to 5. Each retry is an extra attempt. Test Settings
Run
One execution of a pipeline, with a run ID, a status and a report. Schedule building blocks
Run Environment
Where a pipeline executes: a SmartestQA server, your agents, or a mobile device farm. Run Environment
Run list
What a pipeline runs, in order: test cases or requirements, plus sub-pipelines. Add content
Run on Server
Runs the test headless on a SmartestQA server. Nothing to install, but it reaches only apps open to the internet. Run on Server
Run on your Agent
Runs the test on one of your test agents, inside your network, where you can also watch the browser. Run on your agent
Runner
The part of SmartestQA that executes a test case step by step, retries and heals. It works on the Smartest Server and inside your agent. Where tests run

S#

Sample project
A finished example project created once per company. It does not count against the licence. Sample project
Save & Execute
Saves the pipeline and starts a run at once. Run a pipeline now
Scenario
In the Analyze wizard, a test case the AI proposes, created in Automate when you click Create. In InSight, the word for a test case. Analyze building blocks
Scheduled Runs
The Schedule tab that lists pipelines with a Period other than Real-time and their next start time. Scheduled Runs
Screenshot On Fail
A run setting. On: pipeline runs keep screenshots only for failed steps. Dry runs always take one per step. Test Settings
Script Language
The engine of a WEB project, SELENIUM or PLAYWRIGHT. Both run the same test cases. Script Language
Selenium
One of the two browser engines of WEB projects, with its own browser options. Selenium settings
Send (API)
Sends a request now from the SmartestQA server and shows the response. Nothing is saved, and it waits at most 30 seconds. What Send sends
Service name
The short key deploys of a repository appear under, such as web or api. Unique in the project. Repository fields
Setup guide (Setup n/3)
The top-bar checklist of three steps: Connect, Teach your app, Run & report. You tick it yourself; it is kept in your browser. Setup guide
Setup Roadmap
The eight-step progress of a project in Discover, calculated by the server from what exists in the project. Setup Roadmap
Setup and Teardown
Pipeline phases that run once at the start and once at the end. If Setup fails, the tests are skipped; Teardown always runs. Phases
Severity and Status (requirement)
Severity: Low, Medium, High or Critical. Status: New, Assigned, Implemented or Closed. You set both. Requirement fields
Signing secret
A random value shared with a webhook receiver so it can check that a request really came from SmartestQA. Verify the signature
Slack, Teams and WhatsApp
Chat integrations that post pipeline results, task alerts and a daily summary. Slack, Teams & WhatsApp
Smartest AI
SmartestQA’s own AI account, the default for new projects. Model and budget are set by SmartestQA. AI Account
Smartest Server
SmartestQA’s own test server, listed first among the agents. Always online, nothing to install. Agent Pool building blocks
SmartTest Device Farm
SmartestQA’s farm of real phones and emulators, available as an add-on to mobile projects. SmartTest Device Farm
Source (audit log)
How a change arrived: the app, an AI assistant through MCP, an API key, or a request without sign-in. Logs building blocks
Start page URL
A named address with one value per environment, where web recordings and runs begin. One is the project’s default. Start page URL
Step
One line of a test case: an action step (click, type, select) or a step type with its own dialog, such as Assertion, API or Python. Step reference
Sub-pipeline
Another pipeline added as one row of a run list. It runs its own run list inside this run. Sub-pipelines
Sub-requirement
One testable condition with an expected result inside a main requirement. Test cases are linked here. Link test cases
Super Admin
SmartestQA’s operator role. It is not a customer role and cannot be assigned by your Admins. Roles & permissions

T#

Take device
Reserves a free Device Farm device for you and opens its screen in the browser. Farm device
Task
One item of work, such as a defect, on the Tasks board. Failed pipeline runs can create them. Tasks
Task Settings
Where Admins add, rename, colour, order and delete task statuses. Statuses are shared by all projects. Task Settings
Task status
A board column: Open, In Progress, Resolved, Closed, plus statuses your company adds. Statuses
Template
A test case used as one step inside others, such as a login. Editing it changes every test that uses it; its badge shows how many. Templates
Test agent
The SmartestQA client on a computer you control. It runs tests there and sends logs, screenshots and video back. Not the same as an autonomous agent. Test Agents
Test case
An ordered list of steps that runs on its own and produces a report when a pipeline runs it. Test cases and folders
Test Health
The InSight tab that shows how much of the application your tests prove, and what to fix first. Test Health
Test Management
The run settings of a project or pipeline: engine, parallel threads, retries, browser, logging, video, timeouts and report e-mails. Test Management
Tested after
The pipeline runs in the same environment that started after a successful deploy and before the next one. Tested-after window
Thread Count
How many test cases of a pipeline run at the same time, from 1 up to your licence’s Max Thread Count. Thread Count
Trial
The 14-day free start of every new company, with no credit card. Create your account
Trigger
A rule that starts a pipeline, or adds a queue item, when a webhook is called, an e-mail arrives or a file lands in a folder. Triggers

U#

Unreachable (pages)
Pages that no action leads to, counted in a chip on the Action List. Unreachable pages
Update map
Builds or refreshes the App Map from your actions. Run it after big recording sessions and renames. Update map
Usage chip
“This action is used in n test cases”, shown when you select a step. Amber means a change reaches other tests. Action usage chip
Usage mode (device)
Whether a farm device is for Automation, Manual testing or Dual use. Manual devices are never picked for automatic runs. Agent Pool building blocks
User (role)
The role for team members: test work in the projects they were given, no company administration. Roles & permissions

V#

Variable
A named value written as {{name}}, provided by an earlier step, a Data Source key, a dataset or a global parameter. A name nobody provides is typed as written. Variables and test data
Verdict
The one-word result of a run on the report summary: Passed, Passed with warnings, Failed or Running. Run verdict
Video
A recording of a test case built from its screenshots after the test case ends. Video
Virtual Machine agent
A shared test agent created with New Host, for example on a build server. Everyone in the company sees it. New agent
Visibility Check
A step that checks whether an element or text is on screen. A failed check marks the step failed but does not stop the test case. Visibility Check

W#

Webhook
Signed JSON events for pipelines, tasks and deploys sent to any HTTP endpoint you choose. Webhooks
Work key
A short name for a kind of open work that an MCP assistant can pick up, such as SCENARIOS or HEAL. Work keys

X#

Xray
A test management integration with Xray Cloud: sync test cases, bring requirements in and write results back. Xray

Z#

Zephyr Scale
A test management integration: sync test cases, bring Jira requirements in and write pipeline results back. Zephyr Scale

Didn’t find your answer?Write to support with your project name and what you see on screen.

info@smartestqa.ai
© 2026 Selenity Technology · SmartestQAsmartestqa.aiThis page as MarkdownChecked against smarttest-fe predev 2a8c79cc (translations/en.json), smarttest-be 633c2a03; definitions taken from the module pages of these docs, which are verified against the code