Everything Testver does, in detail
Each section below covers a single feature in depth — what it is, why it exists, and what you can do with it.
Any framework, premium experience
Bring AI assistance, recording, management, and reporting to every major test framework — without changing how your team writes tests.
Testver wraps your existing framework rather than replacing it. Point it at a Playwright, Cypress, Selenium, Appium, WebdriverIO, Cucumber, REST Assured, PyTest, Puppeteer, or Robot Framework project — web or native mobile — and the entire UI lights up, from intelligent recording to dashboard-grade reporting, while your test code stays exactly where you keep it.
- Zero migration — works with the projects and tests you already have
- Native adapters per framework — tests run on the framework's own runner
- Web and native mobile (Android & iOS via Appium) under one control plane
- Per-framework code styles preserved in AI-generated tests
AI Assistance
Your copilot for Web, Mobile, API, and Reports.
Ask in plain English. The AI runs the right tools, writes the code, and explains the results — whether you're authoring a Playwright spec, driving an Android or iOS app through Appium, validating a REST endpoint, or building an executive dashboard from last week's runs. Full repository context, live tool use, and every tool call permission-gated by default.
- Web — Create and update tests in any major framework, manage them end-to-end, debug failures, and refactor flaky tests. Or scaffold a brand-new framework of your choice from scratch.
- Mobile — Author, run, and debug Appium tests for Android and iOS, native or hybrid apps, with the same smart selectors and fixes as the web.
- API — Testing, validations, automations, and scripting in minutes.
- Reports — Generate dashboards and reports on the fly.
- Streaming responses with permission-gated tool execution
AI Record Agent
Click. Record. Generate. Save LLM costs by up to 80%.
A dream feature for any test automation framework. Walk through your app once. Testver captures every action with smart, self-healing selectors and generates a clean test in your framework's native style. Automate complex scenarios in minutes with test-aware context — cutting LLM costs by up to 80%. Update or extend tests any time, with your existing structure intact.
- AI Annotations — a first-of-its-kind capability. Add complex validations, custom actions, or data extractions by describing them in plain English, no code required.
- Smart, self-healing selectors keep tests resilient as your UI evolves
- Test-aware context cuts LLM token usage by up to 80%
- Inline action editor + replay before generation; update or extend without rewriting
Test Management
Cases, plans, cycles, executions — one connected system.
Generate test cases with AI in minutes, then link them directly to your automation scripts. Organise everything into test plans, cycles, and executions — with coverage analytics and impact analysis at every level. Full JIRA integration keeps requirements, defects, and team workflows perfectly in sync, so every test traces all the way back to the requirement it covers.
- AI-generated test cases from requirements, PDFs, user stories, or Jira issues
- Link cases directly to automation scripts — bidirectional traceability built in
- Test plans, cycles, and executions with coverage analytics and impact analysis
- Full JIRA integration: requirements, defects, statuses, and comments stay in sync
Runs, Schedules & Results
Run, schedule, triage — all in one view.
Run scripts locally or in the cloud — on demand, on a cron schedule, or as part of your delivery pipeline. Watch executions stream live with per-step screencast, trace, video, screenshots, and console output. Every test keeps its full execution history. AI failure analysis clusters errors and suggests owners; AI Fix turns broken tests into one-click patches. Compare any run against a baseline and dispatch reports to Slack or email.
- Run locally or in the cloud — on demand, on a cron schedule, or via CI
- Live streaming execution with per-step trace, video, screenshots, console output
- Full execution history per test, with baseline comparison and run-vs-run diffs
- AI failure analysis + AI Fix — clustered errors, owner suggestions, one-click patches
Auto-Heal — broken locators repair themselves
A fix built from the failure's own page evidence, proven by a re-run before anything is kept.
Development never pauses for the test suite. A developer renames a test id, restructures a component, or ships a redesign — and tests that were green yesterday can no longer find what they click. When a run fails for that reason, Testver reads the page the test actually saw at the moment it broke — the trace, the live browser session, or a fresh probe — works out which element moved, and proposes the exact replacement. It then re-runs every test affected by that locator and keeps the change only if all of them pass. Nothing is ever a blind guess, and nothing is kept unproven.
- Keeps pace with development — repairs come from the failing page's own DOM, never guesswork
- Verify or revert: every fix is re-run before it is kept, and rolled back if it fails
- Suggest mode reviews first with one-click Apply — or let Auto mode keep verified fixes
- Playwright, Selenium, WebdriverIO, Robot Framework and Cypress — local runs or cloud grids
Mobile testing — Android & iOS, built in
Record, generate, and run native mobile tests with Appium — on real cloud devices or your own USB-connected phones.
Switch a project to Mobile mode and the whole Testver experience follows. Record a session by tapping through your app, let the AI generate a clean Appium test in your framework's style, then run it on a cloud device farm or a phone plugged into your own machine. Live device view, smart self-healing selectors, AI annotations, and the same dashboards, scheduling, and reporting you already get for web — now for native apps.
- Native Android & iOS via Appium — no separate tool, the same Testver UI
- Run on cloud device clouds or local USB-connected devices
- AI Record for mobile: tap-through capture → generated test with self-healing locators
- Web / Mobile / Both workspace modes — surfaces adapt to what you test
AI test case generation
Turn requirements, user stories, or PRDs into a structured suite of test cases — in seconds.
Paste a requirement, upload a PDF spec, or link a Jira ticket. Testver's AI distils acceptance criteria into a complete set of positive, negative, and edge-case test cases — properly tagged, prioritized, and linked back to the source. Export to your TMS, attach to a sprint, or hand off to the recorder for automation.
- Generate from text prompts, PDFs, DOCX, OpenAPI specs, or Jira issues
- Positive / negative / edge / boundary coverage out of the box
- Auto-prioritization with risk-based ranking
- Bidirectional traceability — every test case links to its source requirement
User can log in with valid credentials; invalid attempts must show an error message…
Test Explorer — find, filter, run
Every test in your project, organised and runnable from one place — with stats, filters, tags, and one-click execution.
The Test Explorer groups every test by suite with aggregate stats at a glance: total, passed, failed, no-run, and pass rate. Filter by status, framework, or tag, search by name, sort however suits you. Open a test for the full picture — human-readable steps, source code, run history, streaming output while a run is in flight, and links to its test management mappings. Run a single test, a whole suite, or a multi-framework selection — Testver builds the right command for each runner.
- Aggregate stats with status / framework / tag filters + free-text search
- Per-test detail panel: Steps · Source · Runs · live Output · TM Mapping
- BDD-aware — Cucumber feature files with step-definition navigation
- Run a single test, a suite, or a multi-framework selection — export to CSV / Excel
Project Explorer — browse and edit, no tool-switching
A full file tree + Monaco editor inside Testver — tweak a locator, rename a fixture, or read a config without ever opening VS Code.
Project Explorer pairs a collapsible file tree with a VS Code–grade editor (the same Monaco engine), all inside Testver. Browse your entire project, open any file, edit with syntax highlighting and multi-cursor, search and save with the keyboard shortcuts you already know. Other screens deep-link straight into a specific file at a specific line — click a failing step in Results and the Project Explorer opens that line of the test code, ready to fix. No more bouncing between the test runner and your IDE for a five-second edit.
- Resizable two-pane layout: collapsible file tree + Monaco editor
- Multi-tab editing with drag-to-reorder, middle-click close, dirty-state markers
- Familiar shortcuts: Ctrl+F to search, Ctrl+S to save, multi-cursor selection
- Deep links from Results / Test Explorer jump straight to file:line
Code Map — an X-ray of your test codebase
See the structure under your suite: every method, page object, and helper — who calls what, and what nothing calls at all.
Code Map scans your automation project and builds a live structural index: which tests reach which methods, which methods call which, and which are orphaned. Four interlocking views — Inventory, Quality, Graph, and Steps — surface the load-bearing helpers, dead page objects, over-grown methods, and duplicate code, so you can refactor with confidence and keep the suite maintainable as it grows. Nothing here changes your code; it only reveals what is already there.
- Inventory, Quality, Graph, and Steps views over your whole codebase
- Caller / callee graph — find load-bearing helpers and orphaned code
- Quality signals: over-long methods, duplication, unused page objects
- Read-only insight — reveals structure without touching a line
API testing built in
Postman-grade collections with schema validation, scripting, and AI assistance — no extra tool needed.
Author requests, organize them into collections, define environment variables and auth profiles, and run the whole thing in CI. Schema-aware assertions, JSONPath/JMESPath extraction, pre/post scripts, and AI-generated contract tests round out a first-class API testing surface.
- Collections, environments, auth profiles (OAuth, API key, JWT)
- JSONPath/JMESPath extraction + schema-driven assertions
- Pre-request and post-response JS scripts
- AI-generated contract tests from OpenAPI specs
Connectors for the tools you already use
Two-way integrations with Jira, Slack, and email — configure once, use everywhere.
Push defects to Jira when tests fail. Stream run results into Slack. Email scheduled reports to stakeholders. Each connector is a first-class citizen — configure once and it shows up everywhere it's relevant across the product, from notifications on a failed run to the schedule editor to the AI assistant's tool palette.
- Defect tracker: Jira (with bidirectional status + comment sync)
- Notifications: Slack, email (SMTP)
- Webhooks for any system that can POST JSON
- Configure once — connectors surface everywhere they're relevant
Cloud test-runs integration
Dispatch tests to BrowserStack, Sauce Labs, or LambdaTest — from the same UI.
Configure your cloud provider once, then dispatch any test or suite to run on real browsers and devices in the cloud — without changing a line of test code. Run results, video, and trace artifacts stream back into Testver alongside your local runs, so the reporting view is unified regardless of where the test executed.
- Works with BrowserStack, Sauce Labs, and LambdaTest
- Per-project provider routing — pick where each suite runs
- Artifacts (video, screenshots, traces) auto-imported into the report
- Same dashboards and scheduling whether tests run local or cloud
Scheduling that just works
Cron-driven runs and reports — daily smoke suites, weekly regression, monthly executive digests.
Schedule any test, suite, or report on a cron expression. Built-in scheduler runs without an external orchestrator — no Jenkins, no Airflow, no cron-on-some-server. Per-schedule environment overrides, retry policies, and notification rules; full execution history with on-demand "run now" and pause / resume controls.
- Cron syntax with friendly preset shortcuts (daily, weekdays, hourly)
- Per-schedule environment + connector + notification overrides
- Retry policy with exponential backoff and failure-only alerts
- "Run now" override + pause/resume without losing the schedule
Git workflow, integrated
Commit, branch, stash, push, resolve conflicts — without leaving the test suite.
A polished git surface that handles the things QA engineers actually do: stage and commit changes, manage branches, resolve merge conflicts side-by-side, push over SSH or HTTPS with credential prompts, and view per-file blame and history. With AI on hand to write your commit messages.
- Side-by-side merge conflict resolver with per-hunk picks
- SSH passphrase / HTTPS credential prompts (VS-Code-style)
- Per-file history and blame in the explorer
- Stash, amend, force-push protection — sane defaults built in
CI/CD pipelines that gate on real runs
Your pipeline triggers the tests, Testver executes them, and the build waits for the verdict.
Delegated execution turns the usual CI problem inside out: instead of installing browsers, drivers, and a framework on every runner, your pipeline makes one call and the tests execute on your Testver server — where the environment is already right. The build blocks on the outcome, the verdict is pushed back as a commit status, and every CI run lands in Testver's dashboards with its branch, commit, and environment attached.
- Ready-made templates for GitHub Actions, GitLab CI, Jenkins, Azure Pipelines, Bitbucket and CircleCI
- No browsers, drivers, or framework setup on the CI runner — the server already has them
- Gate the build on the result, or fire-and-forget when you don't want CI waiting
- Deploy-to-tests tracing — follow any commit or deployment through to the runs it triggered
Dashboards & scheduled reports
Generate executive-ready dashboards on demand. Email them on a cron. Share via tokenized links.
Compose KPI tiles, charts, and tables into shareable reports — manually or via the AI assistant. Schedule them daily/weekly, dispatch to stakeholders by email, or share a tokenized public link. Pre-templated executive, QA-daily, release-readiness, and trend-analysis reports ship out of the box.
- KPI tiles, line / bar / pie charts, and data tables
- Pre-built executive, QA, release, and trend templates
- Schedule + dispatch via email / Slack
- Tokenized public share links — no sign-in required
Testver cloud AI or local models
One connection to managed frontier AI — or run Ollama / LM Studio locally for free. Self-host everything.
AI on your terms. Connect Testver's managed cloud AI and one connection powers every AI feature with frontier models — no provider accounts, no API keys, no model configuration. Or point the LLM router at a local Ollama or LM Studio model and the entire AI surface costs nothing. The whole stack — UI, API, orchestrator, browser pool, LLM router — runs on hardware you control, with audit logs and per-user permissions for regulated environments.
- Testver managed cloud AI — frontier models, zero configuration, no API keys to manage
- Local models via Ollama or LM Studio — completely free AI, fully offline
- Per-project AI setup, temperature, and tool-policy overrides
- Self-hosted: your data stays with you; RBAC, audit log, multi-user scoping built in
Agent Skills — teach the AI how your team works
Package your conventions as reusable skills, and the assistant follows them everywhere.
Every team has its own way of working — how tests are organised, how they are named, the standards everyone agreed to follow. An Agent Skill captures those standards as reusable instructions the assistant loads when the task calls for it, so what it produces fits your project instead of reading like a generic sample. Testver ships built-in skills too, including one that scaffolds a complete framework from scratch — so a brand-new project starts out structured the way a mature one ends up. The right skill loads for the right job, keeping results consistent whether the AI is writing, refactoring, or debugging.
- Capture "how we work here" once — the assistant applies it every time
- Built-in skills included — scaffold a complete framework from scratch, properly structured
- Scoped loading: only the relevant skill applies, so prompts stay focused
- Framework-agnostic — the same skill guides Playwright, Cypress, Selenium, and the rest
export class CheckoutPage {
readonly promoInput = this.page.getByTestId('promo');
async applyPromo(code: string) { … }
}MCP servers — extend the AI with your own tools
Connect the assistant to databases, browsers, and internal systems over the open Model Context Protocol.
Testver speaks MCP, the open standard for giving AI assistants real tools. Point it at any MCP server — a database, an internal API, a browser automation server, or one your team wrote — and those tools join the assistant's palette inside your testing workflows. No Testver-specific plugin to build, and every call goes through the same permission gate as the built-in tools.
- Any MCP server, third-party or your own — nothing Testver-specific to write
- External tools sit alongside the built-ins in the assistant's palette
- Permission-gated: extended tools are approved exactly like native ones
- Reach the systems only your network can see — seed data, query state, verify downstream
Intelligent failure analysis
Group, triage, and explain failures automatically — with AI-generated root-cause summaries.
When tests fail at scale, Testver clusters failures by signature and produces a plain-English summary of what changed and why each test likely broke. Linked commits, stack traces, and recent UI diffs all surface in one place — cutting triage from hours to minutes.
- Failure clustering by stack signature
- AI-summarised root cause with linked commits and UI diffs
- Trace + video + screenshots for every failure, side-by-side
- One-click defect creation in your tracker (Jira built in)
Submit button selector renamed from .btn-submit to [data-testid="submit"].
User management & access control
Roles, sessions, and a full audit trail — built in, not bolted on.
Multi-user, multi-tenant ready out of the box. Three default roles — Admin, Tester, and Viewer — gate every meaningful action across the product, with per-resource permissions for finer control. Admins create users, change roles, and revoke active sessions from the same UI; every sensitive event flows into an audit log queryable by user, action, or time range. Secure login uses bcrypt-hashed passwords and JWT sessions you can invalidate at any time.
- Three default roles — Admin, Tester, Viewer — with per-resource permissions
- Admin user management: create, deactivate, change role, revoke active sessions
- Audit log captures every login, role change, credential prompt, and privileged action
- Secure login: bcrypt-hashed passwords + JWT sessions, revocable from the admin UI
Testver Cloud — your own always-on Testver
The full application on a private server we look after, at your own address — awake at any hour, from any device.
Testver on a laptop stops being enough the moment a schedule needs to fire overnight or a second person needs the same test history. Testver Cloud runs the same application on a private machine we build, patch, back up and repair — at your-team.cloud.testver.com, with automatic HTTPS and continuous backups mirrored to a second provider. Everyone signs in to one workspace, and nobody installs anything. Included with the Power and Team plans — one workspace per plan, ready about a minute after you create it from your dashboard.
- Your own subdomain, running 24/7 — schedules fire overnight and at weekends
- One private machine per workspace, never shared — a team plan is one workspace, managed by the owner and one platform manager
- Backed up on a schedule, mirrored off-site, and self-healing when a check fails
- Same Testver, same Git repository — move on or off without touching a test
Built, patched, backed up and repaired for you — nothing to install
Testver Grid — cloud browsers, built in
Run your suite on Chromium, Firefox and WebKit in the cloud — no vendor account, no access keys, no test-code changes.
A browser fleet that is part of Testver rather than a third party bolted on. Pick Testver Grid next to Run on, choose a browser, and your unchanged Selenium or Playwright suite executes against the fleet — with the connection supplied at run time, so nothing vendor-specific enters your config. Session video, an ordered command log, and a verdict proving the run really happened in the cloud land in the same report you already read. Included with Pro, Power and Team plans — a team shares one pool, sized by the plan.
- Chromium, Firefox and WebKit on machines that are not yours — Pro 1, Power 2, Team 3 browsers at once
- No grid-vendor account and no keys in CI — Testver vends a short-lived token itself
- Video, command log and live view for every session, in your normal results
- Reaches the app on your localhost — every hop outbound, nothing newly exposed
No vendor account, no access keys, not a line of test code changed
Every feature above wraps your framework — it never rewrites it
Nothing on this page asks you to migrate your suite, restructure your project, or import a Testver SDK into your tests. Testver adds the intelligence around your tests, not inside them — so the suite you have today keeps running exactly as it does today, with or without us.
No Testver imports, wrappers, or annotations. Your specs stay plain framework tests.
Point Testver at the project you already have. Same files, same runner, same command.
Turn Testver off tomorrow and your suite runs from your terminal exactly as it does today.
The same holds for a framework Testver builds for you from scratch: a standard project in the framework you chose, with no Testver runtime inside it — yours to run, commit, and take anywhere from day one.
Keep your framework. Add AI and a premium toolset.
Set up in minutes. Self-host on your own hardware. Describe a test in plain English and get working code in your own framework — in a fraction of the time.