Skip to main content

Glossary

Key terms used throughout TestFlair and this documentation.

BRD​

Business Requirements Document — the input files or text that describe what your application should do. TestFlair's AI reads your BRD to generate user stories, test cases, and scripts. Supported formats are PDF, Word (.doc/.docx), Rich Text (.rtf), plain text (.txt), and spreadsheets (.xlsx/.csv) — or you can paste requirements directly.

Flairs​

TestFlair's platform credits. Every AI generation (user stories, test cases, test scripts, test data) and the AI steps of a Jira import or export consume Flairs; running tests does not. Track your balance under Settings → Usage — note that the per-feature breakdown there charts attachments, user stories, test cases, and test scripts only; the remaining operations count against your total without a separate bar. Free-tier accounts get a starting allowance; paid plans include higher limits and you can top up at any time.

User Story​

A feature described from an end-user perspective: "As a [user], I want [action] so that [benefit]." User stories are the bridge between raw requirements and concrete test cases. See User Stories.

Test Case​

A set of preconditions, steps, and expected results that verify a feature works correctly. Each test case has a type — Positive, Negative, or Edge — and can be generated by AI or created manually. See Test Cases.

Test Case Type​

A categorization that signals what a test verifies and guides script generation:

  • Positive — the expected, happy-path behavior.
  • Negative — error handling and invalid inputs.
  • Edge — boundary and corner cases.

Test Data​

The input values used by a test step (usernames, amounts, dates). Test data can be entered manually, reused from the project library, or AI-generated. See Test Data.

Test Script​

Executable automation generated from a test case, in Java or Python, using either the Selenium or Playwright framework. See Test Scripts.

Test Run​

A named, rerunnable collection of test cases, executed as a set. A run has one engine and one language, carries its own copy of the execution settings, and never "fails" — it goes Not Started → Running → Completed regardless of how many cases inside it failed. See Test Runs.

Session​

One press of Start or Rerun on a test run. Each session records its own pass/fail counts and progress; previous sessions are kept. A session's Failed status means the session could not run at all — it does not mean some cases failed, which still Completes.

Blocked​

A test case in a run that could not be executed at all, for one of three reasons: it has no script, its script is not ready yet, or the script was built for the other engine. Blocked cases still count toward the run's total, so a run containing them cannot reach 100% progress. See Test Runs.

Flaky​

A test case in a run that ultimately passed but failed at least once along the way. Usually a timing assumption in the script or genuine instability in the application — worth investigating rather than treating as a pass.

API Test​

A test that exercises an application's endpoints directly, without a browser. Generated from your user stories against an imported OpenAPI specification, and refined with variables, assertions and pre/post-request scripts. See API Tests.

OpenAPI Spec​

A machine-readable description of an HTTP API. TestFlair imports one to learn your endpoints rather than asking you to enter them by hand. Specs are versioned, each with its own label and base URL. See API Tests.

TestFlair AI​

The assistant built into the platform. You describe what you want in plain language and it works against your real projects, asking for approval before it changes anything. See TestFlair AI.

Approval Card​

The prompt TestFlair AI shows before it creates, updates or deletes anything. You Accept, Edit or Reject it; nothing is written until you decide. Deletions offer Accept or Reject only. See Approvals.

MCP (Model Context Protocol)​

The open protocol TestFlair speaks so the AI assistant in your editor can drive the platform directly, without switching to the web app. See Connect Your Editor.

Selenium​

A widely used browser-automation framework. TestFlair generates Selenium-based scripts you can run from the platform or download.

Playwright​

A modern browser-automation framework. TestFlair can generate Playwright scripts as an alternative to Selenium.

POM (Page Object Model)​

A design pattern where page-specific elements and actions are separated into dedicated page classes, keeping test logic clean and maintainable. TestFlair can generate scripts in POM style.

Tunnel​

A secure connection (Cloudflare Tunnel or ngrok) that lets TestFlair's cloud browser infrastructure reach an application running locally or behind a firewall. See Testing Local Applications.

Live Browser Viewing​

A streamed view of the actual browser session as your test runs, so you can watch interactions happen in real time. See Live Browser Viewing.

Jira Integration​

A two-way connection to Jira Cloud or self-hosted Jira for importing issues as user stories and bugs, and exporting test artifacts. See Jira Integration.

CI Token​

A per-connection secret that lets a GitLab CI or GitHub Actions pipeline start TestFlair runs. Shown once when generated, stored as TESTFLAIR_CI_TOKEN in your CI, and sent in the X-TestFlair-CI-Token header. See GitLab & GitHub.

Zephyr​

Zephyr Scale — a test-management app for Jira. When enabled, TestFlair can export test cases to Zephyr. See Jira Zephyr.

Beta​

A capability that is still rolling out and is enabled per account or workspace. Features marked Beta in this documentation may not be visible to everyone yet — contact your workspace admin or [email protected] for access.