Glossary
Complete glossary of QA testing terms used in Qualflare. Definitions for workspaces, projects, test cases, launches, defects, and more.
Last updated
This glossary defines the terms Qualflare uses in the app and in these docs, from access tokens to workspaces, and links each term to the page that covers it. Quick reference for key terms and concepts in Qualflare. For general QA and testing definitions — flaky tests, test observability, test management, and more — see the Qualflare QA glossary.
A
Access Token
A project-scoped API key (qfp_…) for CI, scripts and AI agents that can't sign in. The CLI and MCP clients normally sign in instead, as a Connected app. Access tokens are created and managed in Project Settings → Access Tokens. See Access Tokens.
AI Insights
The project page that lists AI recommendations about your test cases, such as flaky or redundant tests. Open it from AI Insights in the project sidebar. See AI Insights.
AI Usage Limits
Separate workspace allowances for AI features, not a shared AI-credit pool or an allowance per seat. Report uploads have their own quota. Quo messages have a monthly limit. Launch analysis, case analysis, case suggestions, coverage analysis and step generation each have a monthly allowance set by the plan, and some plans have no cap on launch analysis. Failure clusters are capped by how many a workspace holds at once, and some plans have no cap on failure clusters. See AI usage limits.
Allow changes
The tick box on the approve page that lets a connected app create and edit cases, suites and defects; unticked, the app can only read. It is the sign-in counterpart of Allow writes. See Connect with sign-in.
Allow writes
The Create Access Token option that lets a token also create and edit test cases, steps, defects and suite coverage over the API and MCP. Uploading test results doesn't need it, and it can't be turned on or off once the token exists. See Creating a token.
Audit Log
A workspace-level record of every significant action taken by team members or the system. Used for compliance and security review. See Audit Logs.
Automated Testing
The practice of using software to execute tests automatically, without manual intervention. Results can be uploaded to Qualflare via the CLI. See Run tests: manual and automated.
B
Browser sign-in
Signing in to Qualflare from another app without pasting a token (OAuth): qf login <identifier> and MCP clients such as Claude and Lovable open app.qualflare.com, where you pick a project and approve. Only members with Editor access or higher can approve. See Authenticate with qf login and Connect a coding agent (MCP).
C
Case Run
An individual test result within a Launch. Represents the outcome (pass, fail, skip) of executing a single Test Case. See Case Runs.
CI/CD
Continuous Integration / Continuous Deployment. The practice of automatically building, testing, and deploying code. Qualflare integrates with CI/CD pipelines via the CLI. See CI/CD Integration.
Claude Code plugin
The Qualflare plugin for Claude Code. From the chat, it generates tests, runs your suite, uploads the results and fixes failures. See Claude Code plugin.
CLI
Command Line Interface. The qf command-line tool that uploads test results from 27 frameworks and 4 generic formats to Qualflare. See CLI Tool. The Claude Code plugin runs qf for you from Claude Code.
Connected app
An app, such as Claude, Lovable or the Qualflare CLI, that a member approved for one project through browser sign-in. It is listed under Connected apps in Project Settings > Access Tokens, where it can be revoked. See Manage and disconnect connected apps.
Cluster
A group of test failures identified as similar by Qualflare's AI. Also called a Failure Cluster. Clusters help identify recurring root causes across multiple test runs. See Failure Clusters.
Coverage Score
An AI-generated score (0–100%) representing how thoroughly a test suite covers different test types, priorities, quality dimensions, and scenarios. Generated by the Quo Agent.
CTRF
The Common Test Report Format, an open JSON format for test results. The CLI reads it with --format ctrf. See Supported formats.
D
Default Token
The access token named Default that a new project may start with, so results can be uploaded before anyone opens Project Settings. It has Read & upload access, without Allow writes, and expires after the default 30 days. See Creating a token.
Defect
A bug or issue found during testing. Defects can be linked to failed test cases and tracked through their resolution lifecycle. See Defects.
E
Environment
The system where tests run (dev, staging, production). Test results in Qualflare can be tagged with environment information. See Environments.
Environment Identifier
An environment's Identifier, such as staging: the value that uploads, the CLI and the API use to name it. It is unique in its project and case-sensitive, and an upload that names one the project doesn't have is refused. See Use environments from the CLI.
F
Failure Cluster
Another name for a Cluster: a group of test failures identified as similar by Qualflare's AI. See Failure Clusters.
Flaky Run
One case run that passed after a failed attempt in the same launch, for example on a retry. Its badge shows the retry count and the Overview's Flake Rate counts it; a test case whose result flips across launches is a Flaky Test instead. See Flaky runs and flaky cases.
Flaky Score
The percentage of a test case's recent results, one per launch, that are the less common result, pass or fail, so it is never above 50%. Qualflare flags the case as a Flaky Test when it has enough results and the score reaches the project's Minority Threshold. See How Qualflare flags a flaky test case.
Flaky Test
A test case whose result flips between pass and fail across recent launches. Qualflare flags a case as flaky when it has enough recent results, both passes and failures appear, and the less common result reaches the project's Minority Threshold. A single case run that passed only after a retry is also marked flaky. Detection is tuned per project under Settings → Quality. See Quality Settings and Detect and fix flaky tests.
I
Idempotency-Key
An optional HTTP header on an upload to POST /api/v1/collect or POST /api/v1/collect/ctrf, with a value made once per upload and sent again on every retry. A retry with the same key returns the launch the first request created and adds nothing, so it isn't counted against the Launch Quota twice; qf <identifier> collect sends one for you. See Retry safely with Idempotency-Key.
Identifier (CLI)
The local name that qf login <identifier> saves a sign-in or project access token under, and the first argument of every project command, as in qf <identifier> collect. The approved project or token, not this name, decides which project receives your results, so the name doesn't have to match the Project Identifier. See Identifier rules.
J
JUnit XML
An XML test report format that many test runners and CI tools write, such as pytest with --junitxml. The CLI reads it with --format junit. See Supported formats.
L
Launch
A test execution event that captures results from running tests. Contains pass/fail summary, individual case runs, and analytics. Can be created manually or via CLI upload. See Launches.
Launch Quota
The number of launches a workspace can create each month, set by its plan and listed on Plan & Billing as Test reports (launches). Each run counts once, whether it is an upload, a browser journey run or a launch you start in the app, and a retry that sends the same Idempotency-Key counts once. See Plans and limits.
M
Manual Testing
Testing performed by humans, as opposed to automated scripts. Qualflare supports tracking both manual and automated test execution. See Run tests: manual and automated.
MCP server
The Qualflare endpoint at https://api.qualflare.com/mcp that AI coding agents such as Claude Code and Cursor call to read and change project data. Clients connect with browser sign-in, or with a project access token if they can't sign in. Changes need Allow changes or a token created with Allow writes. See MCP server reference.
Milestone
A release, sprint or other checkpoint that collects the launches run for it and shows how their tests did. See Milestones.
N
Native Reporter
A Qualflare package for your test framework, such as @qualflare/playwright or qualflare-pytest, that records each run's results with steps, per-attempt retries and attachments in report files, without a network call. qf <identifier> collect then uploads those files as one launch. See Reporters.
O
Open Source Plan
A free Qualflare plan for public, OSI-licensed GitHub repositories, with a public report page and a README badge or banner. Activated per-workspace by linking an eligible repository. See Open Source and its limits on Plans and limits.
P
Pass Rate
The percentage of tests that passed in a launch or across multiple launches. Used to track quality trends over time. See Pass Rate Calculation.
Plan & Billing
The workspace page where the Owner sees usage against the plan's limits, changes the plan, and opens the Stripe billing portal with Manage Billing. See Changing plans.
Project
A container within a Workspace that organizes test cases, launches, and results around a specific product or component. See Projects.
Project Identifier
A project's permanent URL slug, as in app.qualflare.com/project/<identifier>, set when the project is created. It isn't the <identifier> in qf <identifier> collect, which is a name you choose for a login saved on your machine. See Create and manage projects.
Public report
A page at reports.qualflare.com that shows a project's launches to anyone, without signing in. A project in an Open Source workspace publishes one when its Public report toggle is turned on. See Publishing a report.
Q
Qualflare JSON
The JSON report format that every Qualflare native reporter except qualflare_flutter writes, one report per file in a results folder. The CLI reads it with --format qualflare-json. See Supported formats.
Quality Targets
Per-project targets, such as a target success rate, that the project's Overview dashboard compares its results with. Set them in Project Settings > Quality; an empty target or 0 means none. See Quality targets.
Quo Agent
Qualflare's AI assistant (also called "Quo"). Used for coverage analysis, test case generation, step generation, and answering questions about Qualflare from its documentation. See Quo Agent. The AI filter on Explore also uses Quo messages.
Quo Fast and Quo Pro
The two models you can choose for a Quo conversation, where the workspace offers model tiers; otherwise the picker shows a single Quo model. Quo Fast is the default, and a plan without Quo Pro shows it locked in the model picker. See Features by plan.
R
Recent deliveries
The table below the integration cards on a project's Integrations page that lists every attempt to send an integration event, with the provider's error when one fails. Each attempt is Delivered, Failed (a temporary error that Qualflare may retry) or Dropped (a permanent error that is never retried). See Recent deliveries.
Reconnect
A badge on a workspace integration's card: Qualflare can no longer deliver to the provider, for example because the app was removed on the provider's side, so it turned the integration off and kept each project's settings. For GitLab and Linear the fix is to remove the integration and install it again, which deletes those settings. See Remove an integration.
Role
The access a person has in a workspace or a project. Qualflare has five roles, from least to most access: Guest, Viewer, Editor, Maintainer and Owner, and in a project you get the higher of your workspace role and your project role. See Roles and permissions.
Run ID
The ID that most native reporters write into each report as metadata.runId; in CI it is the provider's run ID, which every shard of the run shares. When a results folder holds reports with different run IDs, qf <identifier> collect uploads the newest run and any report without a run ID, unless you pass --allow-mixed-runs. See Run ID.
S
Saved Query
A named, reusable filter combination on the Explore page. Saved queries can be loaded to instantly apply a set of filters to the case table. See Explore Cases and saved queries.
SCIM
SCIM 2.0 provisioning: your identity provider creates, updates and removes Qualflare workspace members to match your directory. It needs a working single sign-on connection first. See SCIM Provisioning.
Seat
One member's place in a workspace on a paid plan. A paid workspace can have as many members as the seats it bought, up to the plan's maximum. See Plans and limits.
Shard
One part of a test run that is split across parallel CI jobs. Pass one result file per shard with --shard, and qf <identifier> collect uploads them as a single launch. See Sharded runs.
Shared Steps
Reusable test steps that can be referenced by multiple test cases. Examples: login procedures, setup steps, common actions. See Shared Steps.
Single Sign-On (SSO)
Signing in to Qualflare through your company's SAML 2.0 identity provider. The workspace Owner connects it on the Single Sign-On page. See Single Sign-On.
T
Team
A group of project members. Suites and defects can be assigned to teams for filtering and responsibility tracking. See Teams.
Test Case
An individual test that defines what to test, steps to execute, and expected results. The fundamental building block of testing in Qualflare. See Test cases: fields, states and steps.
Test Plan
A saved selection of test cases, picked one by one or a whole suite at a time. You run a plan by hand: Create Run on the Test Plans page opens the Test Wizard. Plans have no schedule, assignees or status. See Test Plans.
Test Suite
An organized collection of test cases grouped for management and execution. Examples: "Authentication Tests", "Smoke Tests", "Regression Suite". See Test suites and Browse.
Test Wizard
The page where you run test cases by hand. Open it with Create Run on a test plan, or with Start Launch on the Launches page. It records one verdict per case (Passed, Failed, Blocked or Skipped) and saves the results as a manual launch. Blocked is saved as Aborted. See Run manual tests with the Test Wizard.
Trial
A 14-day free trial of the Core or Scale plan. It starts when you sign up from the plan's link, needs no credit card, and moves the workspace to Starter if you don't subscribe before it ends. See Free trial.
Two-Factor Authentication (2FA)
An account security feature that requires a second verification step (usually a code from an authenticator app) when logging in. See Two-Factor Authentication.
V
Verified Domain
An email domain the workspace has proved it owns by publishing a DNS TXT record. Members whose email address is at a verified domain sign in through your identity provider. See Verified Domains.
W
Webhook
An HTTP callback that Qualflare sends to a URL you set: launch.completed when test results are uploaded with the CLI, a reporter or the API, and defect.created when a defect is created on the Defects page, through the API or MCP, or for a flaky test. Configured per project under Settings → Webhooks. See Webhooks.
Workspace
The top-level container in Qualflare where teams collaborate. Contains projects, manages team members, and controls access. See Workspaces.
See Also
- Concepts - Detailed explanations of core concepts
- Quick Start - Get started with Qualflare
- FAQ - Frequently asked questions
- Analyze launch results - Read a launch's success rate, failures and run history
- Share and export results - Public reports, Slack, webhooks and launches as JSON or CTRF
- Upload results with qf collect - Upload test results from CI or your machine
- Search, filters and shortcuts - Find test cases, launches and defects
Sign-in and account problems
Fix Qualflare sign-in problems: single sign-on errors, two-factor setup at sign-in, locked or deactivated accounts and access errors, with each exact message.
What's new
Where Qualflare announces product and CLI releases, how to check your qf and Claude Code plugin versions, and how these docs mark the version a feature needs.