Turn observation into a structured step
Inspect the rendered page, understand the element’s role and state, choose a compatible action, and carry the selected locator and alternatives into the flow.
Windows desktop · Selenium authoring
Browser automation built from evidence, not guesswork.
WebElemate Studio helps testers and developers inspect real page elements, choose explainable locators, build or record repeatable flows, control waits and recovery, run with actionable diagnostics, and export readable Selenium code for the framework they already use.
3 browser families3 export languages4 data formats0 AI required
Designed to work with your existing stack
ChromeEdgeFirefoxSelenium 4.x targetsC# · Java · Python codeFramework-specific assertions
Why WebElemate
Turn a real browser journey into clear, reusable automation.
WebElemate keeps the inspected element, chosen locator, action, run result, and generated code together. Anyone reviewing the flow can see what the test does, how it finds each element, and where a failure occurred.
Replace fragile selectors with explainable choices
Locator Intelligence creates and validates several candidates, then compares uniqueness, readability, and likely stability. Selectors built from generated classes, session values, or other unstable page details are clearly downgraded.
Bring manual and recorded work into one flow
Add a precise step by hand or record a browser journey at the selected insertion point. Both become editable rows in the ordered Automation grid.
Make browser context explicit
Navigation, opening or switching tabs and windows, and entering frames become visible steps. Studio remembers each tab or window by its order, waits for it to exist during replay, and then switches to it without relying on a temporary browser-session ID.
Catch problems before the browser moves
Preflight checks the run range, required parameters, locators, waits, data columns, row availability, browser state, and context settings before execution.
Debug the step that actually failed
Run one draft action, step through the grid, execute a range, or repeat data-driven runs. Results identify the step, outcome, timing, evidence, and practical next checks.
Generate code that fits the destination team
Choose plain Selenium or a supported test-framework profile for C#, Java, or Python. Framework-specific assertions and dependencies are generated for the selected target.
Keep AI optional and reviewable
The deterministic product works without AI. When Copilot is configured, users choose the provider, the exact context to send, and whether to apply any proposed change.
Product capabilities
Everything needed to move from a live page to reviewed automation.
The main workflow moves through Inspect, Automation, Code, and Results. Recording, test data, preflight checks, browser and version guidance, workspace recovery, and optional AI review support that workflow when needed.
01
INSPECT
Understand the element before automating it.
See identity, type, text or value, page context, relevant attributes, and live state such as visible, enabled, selected, or checked. Accessibility details include the element role and supported ARIA fields such as label, labelled-by, described-by, expanded, selected, checked, current, controls, and popup behavior. Compatible actions and required inputs appear beside the element.
- Live page selection
- Manual Find & select
- Action-aware parameter fields
- Single-step code and execution
02
LOCATORS
Keep alternatives, not just one selector.
Compare active and recommended candidates, use JS Test to query supported locators in the live page without triggering the action, and retain alternatives for later repair.
03
AUTOMATION
Edit the whole journey in one place.
Insert, record, reorder, duplicate, move, and run the visible sequence without rebuilding it elsewhere. Mark a row No run when playback should skip it while code generation, reusable handoff, and AI review continue to include it.
04
DATA & PREFLIGHT
Repeat runs with confidence.
Use CSV, JSON, XLSX, or XLSM data. Validate required columns, row counts, locators, parameters, and browser context before playback starts.
05
CODE
Export readable source for the stack you own.
Generate C#, Java, or Python with supported assertion frameworks. Review and edit the source, then preview the exact destination before saving.
06
RESULTS
Know what happened and what to check next.
See the run and step that executed, duration, PASS, FAIL, or amber HANDLED outcome, captured return values, data source, and whether the WebDriver browser was kept open or closed between runs. Failures include categorized repair guidance and redacted technical details.
07
WORKSPACES
Resume authoring without rebuilding the flow.
A saved workspace restores ordered steps, waits, locator choices, values and data mappings, run settings, browser and code-profile choices, editor content, and relevant layout state. Live browser handles, Results reports, AI chats, and secrets are not stored in the portable workspace. Local recovery provides an additional snapshot for unsaved authoring work.
08
WAIT, SCROLL & RECOVERY
Make timing and failure behavior visible.
Build an ordered wait plan, optionally center and verify the target before interaction, and choose typed exception handling that either stops the flow or continues with an amber HANDLED result.
09
KEYBOARD
Model meaningful keys—not just text.
Add Enter, Tab, Shift+Tab, Escape, arrows, function keys, modifiers, or another supported shortcut to a selected element or to the browser's active element.
10
REUSABLE CODE
Hand developers a compact copy kit.
Generate imports, shared helpers, named locators, referenced data, and an ordered Step 1…N call block without creating one duplicate method per step. Studio never executes this read-only handoff.
11
AI COPILOT
Ask for review without surrendering control.
Choose redacted context, preview the payload, request analysis or code review, and transfer only explicitly selected C# step bodies. Java, Python, and other translations remain review-and-export artifacts.
One connected path
From first click to maintainable code.
A clear path for first-time users, with enough evidence and control for experienced automation engineers.
Open
Choose an installed Chrome, Edge, or Firefox browser and navigate to an authorized test environment.
Inspect
Select an element, review its semantics and locator candidates, then choose the action that matches the control.
Build or record
Add the step exactly where it belongs, or capture the rest of the journey into the same Automation grid.
Control timing and recovery
Arrange condition-based waits, enable scrolling when a target may be obscured, and define what each expected exception should do.
Validate and run
Set data, range, and repeat count. Preflight checks the visible flow before the browser begins, then Results separates passed, failed, and handled outcomes.
Review and export
Generate ordinary flow code or a reusable developer copy kit, optionally review selected C# step bodies with AI, then safely save C#, Java, or Python for the destination project.
Value for the team
One workflow that the whole team can understand and use.
People are not limited to one role in WebElemate. Anyone can inspect, build, record, run, debug, review, or export a flow according to the work they need to complete.
The inspected element, locator choices, ordered steps, waits, data, Results, and generated code remain easy to review together.
Live locator validation, JS Test, preflight checks, and single-step execution catch common problems before a full run or code handoff.
Readable C#, Java, or Python code follows the selected framework’s structure, dependencies, setup, cleanup, and assertion style.
Teams choose browser settings, data, run range, export destination, update behavior, AI provider, shared context, and whether to apply proposed changes.
Guided product tutorial
Learn each workflow with the screen, the steps, and the first checks.
Product screenshots now sit beside the feature they explain. Each guide covers the controls, likely challenges, and what to inspect before changing the flow.
Open the complete documentation tutorial →Confirm the browser context, element evidence, step order, policies, and Results before changing code.
Check readiness, locator, frame or tab context, interception, and page state before increasing a timeout.
Generated and AI-assisted source remains developer-owned code and must fit the destination repository.
FIRST RUN
Build and verify a useful flow in minutes
- Navigate to the test page.Enter an authorized HTTP or HTTPS address and choose Navigate. If the browser session ended, the button changes to Restart & navigate.
- Choose Inspect.Click the element you want in the automated browser. Studio loads its details, compatible actions, and locator evidence.
- Prepare the action.Choose the action, enter any required value, and select the locator you want the step to use.
- Test or add the step.Run step validates the draft immediately. Add step places it into the shared Automation flow; testing first is optional.
- Build the journey.Set Insert at, start recording, complete the remaining browser actions, and stop recording. Manual and recorded steps appear together.
- Run and review.Choose From, To, Runs, and any data settings. Run flow performs preflight, executes the range, and opens Results.
- Generate the destination source.Select a Framework profile, choose Code, review the output, and use Save to preview the file action before writing.
- Save the workspace.Use File > Save workspace as… to keep the authoring document for review and later work.
Check first if the draft failsConfirm the active Browser Tab, frame or shadow context, locator match, and operation type before increasing a timeout.
Question: does Run step add the row?No. Run step tests the Inspect draft. Add step is the deliberate authoring action.
FILE & WORKSPACE
Save the authoring state, not a live browser session
New workspace · Ctrl+NStarts a clean untitled authoring session after offering to save unsaved work.
Open workspace… · Ctrl+OLoads a versioned .saworkspace.json document. Supported older formats are imported without rewriting the original.
Save workspace · Ctrl+SWrites the current workspace to its existing path.
Save workspace as… · Ctrl+Shift+SChooses a new file path. Use this for a portable copy or a new scenario.
Check for Studio updatesReads the configured release manifest and reports availability. It does not silently install an executable.
Exit · Alt+F4Closes Studio after the unsaved-work decision.
Workspace name and dotThe name beside File is informational. A dot means the current workspace has unsaved changes.
Local recoveryNormal shutdown keeps a private local recovery snapshot of unsaved authoring work. Recovery is helpful, but it is not a substitute for Save or source control.
Workspace contentsSteps, locators, waits, data settings, code profile, generated editors, browser choice, and relevant UI state are portable. WebDriver handles, AI chat, execution results, and secrets are not.
Check first after reopeningA workspace restores authoring state, not the old live browser. Use Navigate or Restart & navigate, then re-establish any required page state.
Question: is recovery a backup?No. Local recovery protects interrupted authoring work; keep deliberate workspace copies and exported code in approved storage or source control.
APPLICATION HEADER
Control the browser, generation target, and product help
Back / Forward / RefreshMoves browser history or reloads the page. The matching browser action is added to Automation.
Browser Tab NBrings the retained browser forward and identifies the active zero-based recording context as a friendly one-based tab number. Its dropdown switches recording focus and records exactly one indexed Switch step.
Address boxAccepts a complete URL and suggests saved addresses while typing. Ordinary website addresses without a scheme are normalized to HTTPS.
Save URLSaves the current address for later suggestions without navigating or adding a step.
Navigate / Restart & navigateUses the current tab, or starts a fresh browser session when Studio is offline.
Maximize on runThe square toggle beside Navigate is on for a new workspace. When checked, flow playback normally maximizes the browser without hiding its tabs, address bar, or Windows taskbar. It does not resize the browser immediately.
BrowserSelects an installed Chrome, Edge, or Firefox family. Studio does not install or uninstall the browser.
Browser statusShows starting, online, busy, or offline state.
Inspect / Stop inspectStarts or stops element selection across the automated tabs and windows.
Framework profileChooses the exported language and optional assertion framework. It does not change Studio playback.
Selenium APIChooses an enabled Selenium 4.x target for generated source and dependency guidance.
Version guideOpens browser, driver, Selenium target, dependency, path, cache, proxy, and compatibility guidance.
Implicit wait msSets the flow-wide WebDriver element lookup wait. Zero disables implicit waiting.
Explicit wait msSeeds the Suggested wait for new eligible rows. A row plan edited and saved by the user remains independent when this header value later changes or becomes zero.
Help / F1Opens searchable, task-oriented guidance inside Studio.
AI Copilot / Disconnect AIOpens the optional Copilot window or removes the active connection and remembered credential.
Update availableAppears only when the approved update manifest reports a newer compatible release.
Tabs and windowsBrowser Tab N is the deterministic context selector. It is refreshed as tabs open or close and stays aligned with recording focus. Browser action remains the place to author explicit Navigate, Open, Switch, Close tab/window, Close browser, history, refresh, and browser-key steps.
Check first when the header says offlineConfirm the browser is installed, the previous session was not closed, and the configured driver path or Selenium Manager policy is valid.
Question: when does Maximize apply?It applies when flow playback starts. It does not immediately resize the browser while authoring.
INSPECT
Understand the target and create the right step
Inspect / Stop inspectChoose Inspect once, then click an element in any automated tab or window. Studio intercepts the selection click instead of sending it to the page.
Selected elementShows identity, classified type, role, accessible name, text or value, state, attributes, page context, and available options where relevant.
ShowFilters operations by category, including actions, state reads, assertions, waits, information, groups, and advanced behavior.
ActionLists only operations compatible with the selected element. A native select, checkbox, file input, text input, and button receive different choices.
Parameter fieldsAppear only when required—for example text, expected value, timeout, option text or value, index, file, or output variable.
CodeOpens a closable single-step source preview. It never replaces the full-flow code workspace.
Run stepExecutes the displayed draft against the live browser and opens Results.
Add stepAdds the current action, values, active locator, and retained alternatives at the selected insertion point. Blue is untested, green passed, and red failed.
Find & selectSearches by visible text or an entered CSS, XPath, ID, name, role, or test identifier, highlights the strongest visible match, and loads it into Inspect.
Copyable valuesDisplayed attributes and select-option values can be copied. Do not put real credentials in a workspace intended for sharing.
Check first when Inspect cannot selectBring the intended Browser Tab forward, close blocking native dialogs, and verify the target is in an accessible frame or open shadow root.
Question: why are some values protected?Credentials and sensitive fields are deliberately unavailable for copying, diagnostics, and shared workspace evidence.
LOCATOR INTELLIGENCE
Compare candidates and keep the reasoning
WebElemate can create ID, name, CSS, XPath, link-text, role/accessibility, stable navigation-path, semantic-parent, and structural candidates. It validates the exact inspected DOM node, checks uniqueness, recognizes unstable tokens and generated page attributes, and scores expected durability.
TypeIdentifies the locator strategy.
LocatorShows the value. Right-click to copy the value alone or with its type.
Score / GradeSummarizes ranking confidence; it is guidance, not a guarantee that a website will never change.
StatusReports live validation, match count, uniqueness, and relevant warnings.
JS TestQueries supported CSS, XPath, ID, name, and link-text strategies in the live document without triggering the target action.
InsightExplains strengths, penalties, stability concerns, and validation findings.
Select a rowClick a candidate to make it active. Blue marks the active locator; green marks the strongest validated recommendation.
Sort a columnClick a column title to sort. Clicking a row selects it; row clicks do not change the sort order.
About JS TestJS Test is a read-only locator check, not JavaScript code export. For supported CSS, XPath, ID, name, and link-text rows, it uses JavaScript in the current document or frame to report matches and whether the exact inspected target was found, without clicking or changing the page. Shadow-root rows continue to use the validated Selenium result.
Strong choicesPrefer stable IDs, names, roles, accessible names, tested data attributes, and semantic scope. Dynamic URLs, generated classes, session tokens, translated prose, and absolute DOM paths are downgraded when a stronger option exists.
Check first when the match count changesRe-inspect the live element, verify the frame or tab, then compare the retained alternatives before manually rewriting a selector.
Question: is the score a guarantee?No. It is evidence-based guidance; the destination team still owns selector maintenance as the site evolves.
AUTOMATION
Edit, arrange, and control the complete flow
CodeGenerates every Automation row, including rows marked No run, into the persistent Flow code workspace.
ReusableBuilds a separate read-only developer copy kit with shared helpers and one ordered Step block.
From / ToSelects an inclusive playback range. To = End continues through the final row.
Run / StopPerforms preflight and executes the selected range. While active, Stop requests cancellation between browser commands.
Insert atSets the exact step number for the next manual or recorded action.
InspectReturns to Inspect and starts selection when a browser is available.
Record / StopCaptures supported browser interactions at the selected insertion point.
Browser actionAdds Navigate, Open new tab/window, Switch or Close tab/window, Close browser, Back, Forward, Refresh, or a browser keyboard action. Navigate/Open accept an optional URL; blank remains blank.
DeleteEnters row-selection mode so several checked steps can be removed in one confirmed action.
Step ◀ / Step ▶Executes the highlighted runnable row and moves the debug cursor to the previous or next runnable row.
Undo / RedoEach click restores exactly one user action, with up to 100 flow-grid actions retained.
Move to… / Up / Down / dragMoves one row or a multi-row selection while preserving relative order.
Copy / DuplicateCopies a review-friendly description or inserts independent copies after the last selected row.
ClearRemoves the entire flow after confirmation and resets From to 1 and To to End.
Editable grid fields
No run appears immediately after #. A checked row is skipped only during flow playback; it remains available to Code, Reusable, AI review, save/load, move, copy, and all other authoring operations. Wait, Scroll to view, Exception, AI step code, Data variable, Data column, Value, Locator name, Selected locator, Locator choices, and Update from Inspect stay with each row. Operation is fixed after creation. The grid scrolls horizontally so every field remains reachable.
Check first when a row runs unexpectedlyConfirm From/To, No run, the selected row, global Wait and Scroll switches, and whether a reviewed AI body is attached.
Question: does Clear remove code files?No. It clears the flow after confirmation and resets the range; it does not delete exported source from disk.
WAIT, SCROLL & RECOVERY
Make every timing and recovery decision reviewable
Explicit wait msSeeds the visible Suggested wait on new eligible steps. Once a row's plan is saved manually, later header changes do not overwrite it.
WaitOpens the ordered plan. Add, remove, reorder, and edit Sleep, readiness, DOM, visibility, clickability, hidden, text, attribute, property, value, or URL conditions.
Suggested waitsA new step may begin with one operation-aware suggestion, such as Clickable for an interaction or Present in DOM for a read. It is ordinary editable data.
Wait ON/OFF titleGlobally includes or omits saved wait plans in both Run and generated code without deleting their configuration.
Scroll to viewResolves the target, centers it, waits for stable geometry, and for pointer actions checks that a sticky header or overlay does not cover the hit point.
Scroll ON/OFF titleGlobally includes or omits checked row scrolling while preserving each row's setting.
ExceptionOpens an ordered list of typed catches for common Selenium and browser-context failures. Put the most specific catch first; Any must be last.
After catchContinue marks the row amber HANDLED and proceeds. Stop reports failure and ends the current flow execution.
Results messageAdds diagnostic text to Results and generated logs. It is never executed as code.
Finally reportAdds an optional always-run cleanup message after the operation or catch outcome.
Execution orderFor each element row, Studio performs enabled Wait items, enabled Scroll to view, the Operation, then matching exception handling. The same visible order is emitted in generated source.
Check first when a click is interceptedLook for a cookie banner, modal, loading layer, or sticky header at the hit point; then use the readiness signal tied to that blocker.
Question: should Wait or Scroll run first?Wait runs first, Scroll to view second, then the operation. This keeps target resolution and readiness ahead of viewport positioning.
RECORDING & CONTEXT
Capture browser work without hiding the result
- Set Insert at.Choose where captured actions belong in the existing flow.
- Choose Record.The control changes to Stop. Supported typing, meaningful keys, selection, checking, clicking, navigation, and history events enter the grid in observed order.
- Move through tabs and windows normally.A blank tab opened with the browser's + button records Open new tab and stays selected; the internal New Tab page is ignored. A real address entered there records a separate Navigate step. Native tab selection or Browser Tab N records an indexed Switch step without forcing Studio back to another tab.
- Close the intended context.Closing one tab records one Close tab/window step and continues in the nearest remaining tab. Closing the browser records one Close browser step rather than several synthetic tab closes.
- Work inside frames when needed.Frame context is recorded for targets inside iframes. Studio returns to the top document only when the next step requires it.
- Choose Stop.The captured rows are already part of the shared flow and can be edited, moved, generated, or run like manual steps.
How replay finds the right tab or windowStudio waits until the recorded number of tabs or windows exists, then switches to the saved position—for example, context 0 for the original tab and context 1 for the next one opened. It does not search every tab by URL, reducing flicker and avoiding accidental navigation.
Check first when a switch is missingConfirm Browser Tab N changed to the tab you selected before performing the next action. Use its dropdown as the deterministic fallback.
Question: should a blank new tab add Navigate?No. Open new tab is recorded first; Navigate is added only after a real URL is entered or another user navigation occurs.
SEMANTIC KEYBOARD ACTIONS
Record or author the key that changes the page
Press key / shortcutFor a selected element, choose the action and enter Enter, Tab, Shift+Tab, Escape, Space, arrows, Home, End, PageUp, PageDown, Delete, F1–F12, or a Control/Alt/Shift shortcut.
Press browser key / shortcutFrom Browser action, send the key to the browser's currently active element when no saved locator should own it.
Plain typingUse Enter text or Append text for ordinary text. Use the keyboard action for navigation keys, semantic submission, function keys, and shortcuts.
Plus keyWrite Plus when the + key is part of a shortcut—for example Control+Plus—so it is not confused with the shortcut separator.
RecordingNormal typing stays one text row. A meaningful key such as Enter becomes its own action; navigation it triggers is not duplicated as an unrelated Navigate row.
Generated codeThe selected key or shortcut is preserved in generated C#, Java, and Python and in the reusable copy kit.
Check first when Enter does not submitConfirm the row is Press key/shortcut with value Enter, the correct target or active element has focus, and the destination page has a semantic submit handler.
Question: when should I use a browser key?Use it when the current active element should receive the shortcut and no saved element locator should own the action.
DATA & PREFLIGHT
Repeat a verified range with external test data
RunsRepeats the selected From/To range from 1 through 10,000. The control is wide enough to show a four-digit value without hiding it.
Browser close On/OffThe unlabeled switch after Runs controls “Close browser after every run”; hover shows the complete meaning. Off retains the WebDriver browser. On closes it after each iteration and starts the next run with a fresh session.
Data fileSelects bounded CSV, JSON, XLSX, or XLSM input. The button changes color when a file is selected; hover shows its path instead of consuming row space with “No data file.”
Clear data fileRemoves the file association without deleting the source file.
Use data fileEnables row substitution. When off, every step uses its hardcoded grid Value even if Data column annotations remain.
Data variableNames reusable generated input or output state. Reusing the same valid name reuses the variable.
Data columnMaps an input to an external-data header. A column on a step that consumes no input is guidance, not a blocker.
What preflight checks
Before the browser moves, preflight validates the requested range, supported operations, required locators, locator types and values, parameters, waits, data file, row count, required columns, tab/window indexes, and settings required by the chosen browser. Blocking rows turn red and do not run partially.
Check first when preflight blocks the runRead the exact row and field named in the message. Verify required data columns and available records before changing the browser flow.
Question: what happens when Use data file is off?The file remains associated, but playback deliberately uses the hardcoded values displayed in the grid.
RUN & RESULTS
Debug at the right level and keep the evidence
Run stepTests the current Inspect draft against the live browser.
Step ◀ / Step ▶Runs the selected grid row and advances the debug cursor without launching an unrelated new flow.
RunRuns the selected range for the requested iterations after preflight succeeds.
ResultsShows run number, range, data source, browser lifecycle, step path, duration, PASS, FAIL, or HANDLED state, returned values, technical evidence, and repair guidance.
Run GuidanceDistinguishes navigation or SPA readiness, locators, frames, shadow roots, dialogs, click interception, stale elements, viewport issues, assertions, uploads, scripts, and browser sessions.
Structural failure snapshotOn Selenium failures, captures bounded page structure, counts, ready state, active-element attributes, and a redacted target summary—not cookies, storage, passwords, form values, or the full DOM.
ClearClears only the visible Results report. It does not remove flow rows, workspace history, code, or files.
Row colorsPassing rows become green, handled failures amber, and failed or invalid rows red; the current selection remains distinct.
Closed browser behaviorSingle-step debugging never silently restarts a closed browser. Use Restart & navigate so the active start page and session are explicit.
Check first after a FAILClassify it first: readiness, locator, frame/tab context, interception, stale element, assertion, script, upload, or browser session.
Question: is HANDLED successful?No. Amber HANDLED means the configured Continue policy absorbed a known failure and moved to the next step.
CODE & EXPORT
Generate source for the destination project—not a demo snippet
Framework profileChoose C# plain, NUnit, xUnit, or MSTest; Java plain, JUnit 5, or TestNG; or Python plain, pytest, or unittest.
Selenium APIChooses the enabled Selenium 4.x source target and matching dependency guidance.
CodeGenerates the complete included Automation flow. The Code tab also keeps a separate closable preview for one Inspect action.
Line numbersEvery populated code surface displays stable one-based line numbers, including reviewed AI code and export previews.
Find / Ctrl+FSearches case-insensitively. Enter or F3 moves forward; Shift+Enter or Shift+F3 moves backward; Escape closes search.
Minimize / Restore Code / ResultsCollapses or restores the right workspace to give the Automation grid more room.
Undo / RedoMoves through up to 100 direct editor states independently in the step preview and full-flow editor.
CopyCopies the source currently visible in the editor.
Save…Choose a destination, then review the exact action and text change. The preview classifies it as Create, Skip, Conflict, or Reject.
Replace approvalAn existing different file requires an explicit Replace selection before Apply export. A stale preview stops before writing.
ClearClears editor text only. It does not delete Automation rows.
Generated / Edited exportGenerated matches the current Automation grid. A direct editor change is an export-only draft until it is transferred as reviewed C# step code.
Important execution modelStudio playback always uses its tested built-in C# Selenium engine. Java and Python profiles control exported source; their compilers and runtimes belong in the destination project or CI system. JavaScript generation is not exposed in the current Studio interface.
Check first before exportCompare each Step marker with its grid row, verify imports and dependency versions, and compile Java or Python inside the destination project.
Question: does editing Flow code change the grid?No. Direct editor text is an export draft. Only an explicitly transferred reviewed C# step body becomes attached to a grid step.
REUSABLE DEVELOPER CODE
Create a compact handoff without changing the flow
- Build or record the complete flow.Reusable code represents the current Automation order, waits, scrolling, context, policies, locators, and step data.
- Select the destination profile.Choose one of the currently exposed C#, Java, or Python framework profiles and the required Selenium target.
- Choose Reusable.Studio opens a separate read-only copy kit; it does not replace Flow code and is never executed by Studio.
- Review the five parts.Imports appear first, then deduplicated helpers, named By locators, only referenced data and policy values, and one flat Step 1…N block of helper calls.
- Search, copy, or save.Use Find or Ctrl+F, then move the useful parts into the appropriate class or module in the destination project.
- Regenerate after changes.A changed row, wait default, policy switch, framework, or Selenium target makes the previous copy kit stale.
Deliberate AI attachmentAfter generation, Copilot exposes Reusable code as a separate context checkbox. It starts unchecked and must be explicitly selected for each request; Privacy preview shows the redacted bounded payload.
Check first when a helper is missingConfirm at least one current step needs that helper. The copy kit intentionally omits unused methods, locators, and data.
Question: why is the Step block flat?It keeps the original execution order easy to copy while shared behavior remains in parameterized reusable helpers above it.
VERSION GUIDE
Understand browser, driver, and Selenium compatibility
Refresh local checkRepeats browser, driver, Selenium target, and optional binding checks.
Reload settingsReloads the unified webelemate.json after a deliberate administrator edit.
Copy dependencyCopies destination-project package or dependency instructions for the selected profile and Selenium target.
Copy startup codeCopies optional startup code for the generated destination project.
Browser downloadOpens the official browser vendor page. WebElemate does not download a browser.
Driver sourceOpens the official driver guidance or source.
Selenium releasesOpens the official Selenium release page.
CloseReturns to Studio.
Two supported operating paths
Zero-setup: leave browser and driver paths unset, use automatic driver mode, and allow Selenium Manager to resolve and cache a matching driver. Managed enterprise: set approved per-browser paths, pin matching drivers, use a prepared offline cache or proxy, and control downloads through policy.
Check first when a driver cannot startCompare browser and driver versions, configured paths, download policy, proxy, cache permissions, and endpoint protection.
Question: does WebElemate install browsers?No. It controls an installed browser and links to official vendor sources when installation or updates are required.
OPTIONAL AI COPILOT
Send only the context you choose and review every proposal
AI CopilotOpens a separate review window. It remains offline until a saved connection successfully tests or completes a request.
ProviderSelects a configured OpenAI API, Azure OpenAI, Anthropic, Google Gemini, Amazon Bedrock, OpenAI-compatible, local/offline, or built-in test connection.
Connections…Opens provider setup, field help, credential behavior, saved choices, and connection testing.
Context checkboxesChoose Element, Locators, Nearby DOM, Steps, Flow code, AI generated code, Reusable code when available, or Last result. Clear anything the provider should not receive.
Privacy previewShows the redacted, size-limited payload before transmission. Nearby DOM is a limited fragment, not the full page.
Verify elementRequests a grounded review of the currently inspected target.
Improve locatorRequests alternatives and rationale from the selected locator evidence.
Diagnose failureReviews the last redacted execution result.
Review generated codeRequests language- and framework-aware code review.
Implement custom codeRequests complete reviewed source for conditions, match-count checks, loops, or other logic the ordinary row fields cannot express.
Transfer step block(s)For C# proposals with exact Step markers, opens a chooser. Row selection previews; the unchecked Use box alone controls which operation bodies are transferred.
Translate with AIFrom Code, prepares a Copilot request from Flow code. It does not send automatically; translations remain review, copy, and export artifacts.
Send / Cancel / Clear chatSends the previewed context and prompt, cancels a supported request, or clears only the visible conversation.
New / Reload JSON / Save / Test / Forget / DeleteCreates, reloads, saves, tests, removes only a credential, or deletes a connection and its remembered credential.
Remember on this deviceStores the credential in Windows Credential Manager. Without it, the credential is kept for the current session only.
Disconnect AIEnds the active session. A session-only credential is cleared; a remembered credential remains protected for reconnection until it is forgotten or the connection is deleted.
A safe custom-step workflow
- Attach only necessary context.Keep Reusable code and unrelated steps unchecked unless the request truly needs them; verify the Privacy preview before Send.
- Ask for one bounded body.Request a complete C# body under the exact
// Step Nmarker and tell the model not to duplicate grid-owned waits, scrolling, or catch wrappers. - Review control flow.Check collection bounds, loop exit, click count, locator reuse, exception behavior, and any values that became hardcoded.
- Transfer deliberately.In Choose reviewed AI step blocks, select a row to preview it and check Use only for each operation body you approve.
- Confirm the attached bodies.Choose Transfer checked step(s), verify the affected rows show AI edited..., and open any marker to review, save, or restore that row.
- Run the smallest range.Regenerate Flow code, run only the affected step or range first, and treat compilation as necessary—not sufficient—evidence.
“For Step 5, use FindElements with the existing locator. If count is greater than 0, click the first displayed and enabled match; otherwise throw NoSuchElementException. Return only the C# operation body. Do not add waits, scrolling, or catches.”
“For Step 8, use a bounded for loop over the existing matches. Click once, set a flag, break immediately, and throw a clear exception if no usable element is found.”
“For Step 11, if the match collection length is more than 0 and the first input is displayed and enabled, clear it and enter the existing step value; otherwise throw a clear Selenium exception.”
Control stays localStudio does not continuously monitor the browser, silently rewrite a flow, or execute source inside Copilot. A reviewed C# operation body runs only later through the normal Automation flow, surrounded by the grid-owned policies.
Reject these patternsUnbounded loops, hidden retries, arbitrary sleeps, credentials, new browser sessions, unsupported network/file/process access, and locator changes without evidence.
Question: does a translated result change Automation?No. Translations remain review, copy, and export artifacts. Only explicitly transferred supported C# Step bodies attach to grid rows.
Verified scope
Clear support boundaries build buyer confidence.
WebElemate is a focused Windows desktop authoring and playback product. It complements the engineering stack around it.
Browser and execution
Windows desktopChromeEdgeFirefoxBuilt-in C# playbackSelenium Manager
Studio controls an installed browser through WebDriver. It does not install the browser or act as a hosted execution farm.
Generated source and assertions
C# plainNUnitxUnitMSTestJava plainJUnit 5TestNGPython plainpytestunittest
The selected profile changes test structure, imports and dependencies, setup and cleanup, assertions, and failure APIs. Generated source must still be reviewed, compiled, tested, and integrated in the destination repository.
Test data
CSVJSONXLSXXLSMNamed variablesColumn mapping
Map inputs to data columns, choose the run range and count, and let preflight verify the file, required columns, and available rows before the browser moves.
Browser context
TabsWindowsFramesOpen shadow rootsRecorded context switches
Opening and switching contexts remain visible operations. Replay waits for the required tab or window position and restores frame context only when a step needs it.
Clear product boundaries
WebElemate focuses on authorized browser-test authoring and playback. Teams still own target permission, destination-code review, security policy, provider terms, and production approval.
Designed for responsible use
Deterministic by default. Optional services stay explicit.
Inspection, recording, playback, diagnostics, and code generation work without an AI connection.
Users select locators, actions, waits, data, ranges, export destinations, and whether to apply proposals.
Diagnostics and AI context apply redaction; credentials use protected Windows storage when remembered.
Administrators can control browser and driver paths, downloads, caches, proxies, host restrictions, AI providers, and update manifests.
Product leadership
“I created WebElemate Studio to make reliable browser automation easier to build, review, and improve—without taking control away from the people who own the tests.”
Dhaval Parikh is the creator and lead architect of WebElemate Studio. The product direction is practical: keep browser evidence visible, make automation decisions understandable, produce code teams can review, and ensure execution, data, diagnostics, exports, and optional AI assistance remain under user control.
Dhaval ParikhCreator & Lead Architect · WebElemate Studio
Frequently asked questions
Important answers before evaluation.
Who is WebElemate Studio for?
It is for anyone responsible for understanding, building, reviewing, or maintaining browser tests—including testers, developers, automation engineers, analysts, domain experts, accessibility specialists, reviewers, and technical leaders. The same person can use any part of the workflow; WebElemate does not restrict features by job title.
Does Studio execute Java or Python inside the product?
No. Playback intentionally uses Studio’s tested built-in C# Selenium engine. Java and Python profiles generate destination source; those runtimes and builds belong in the destination project or CI system.
Does the current Studio export JavaScript?
No. The current user interface exposes C#, Java, and Python profiles. JavaScript-related implementation settings are reserved and should not be treated as a current product promise.
Does WebElemate install a browser or driver?
It does not install or uninstall browsers. For WebDriver, Selenium Manager can resolve and cache a matching driver automatically, or an administrator can configure a pinned driver or prepared offline cache.
Why retain several locators for one step?
Websites change. Keeping validated alternatives preserves the original reasoning and allows repair without repeating the entire authoring journey.
What does JS Test do?
JS Test checks a supported CSS, XPath, ID, name, or link-text locator against the current live document or frame without performing the target action. It reports matches and whether the exact inspected element was found. It is a locator-verification feature, not JavaScript code export, and shadow-root rows use the validated Selenium result instead.
How are tabs and windows handled?
Open and Switch are separate operations. Studio numbers tabs and windows from zero in the order they appear. During replay it waits until the required number of contexts exists, then switches to the saved position instead of depending on a temporary browser-session handle or searching by URL.
What is the difference between Implicit wait and a row Wait plan?
Implicit wait is WebDriver's flow-wide element lookup policy. Explicit wait ms only seeds a visible Suggested condition for new eligible rows. Each row can then keep its own ordered conditions, and a manually saved plan is not overwritten by later header changes.
What happens when an exception policy continues?
The matching row is marked amber HANDLED, its optional diagnostic message is included, and execution continues with the next step. It is not counted or colored as a pass. A Stop rule reports failure and ends that flow execution.
What is Reusable code?
It is a read-only developer copy kit: imports, deduplicated helpers, named locators, referenced data and policy values, and one ordered Step block. Studio never executes it, and it does not replace the ordinary Flow code editor.
Will saving generated code overwrite a file automatically?
No. Save opens a review-first export preview. A different existing file is a Conflict and requires explicit Replace approval. If the file changes after preview, the stale plan stops before writing.
Is AI required?
No. The core product is deterministic and works with AI offline. Copilot is an optional review assistant that receives only the context categories selected by the user after privacy preview and redaction.
Does AI output run automatically?
No. Copilot is a review surface, not a second execution engine. Only explicitly checked supported C# Step bodies can be transferred to their matching Automation rows. They run later through the ordinary flow, surrounded by the grid-owned wait, scroll, and exception policies.
Does WebElemate replace our repository or CI/CD platform?
No. Your repository, test framework, builds, code review, and CI remain the source of truth. WebElemate improves browser-based authoring, validation, debugging, and code handoff.
Does generated code still need review?
Yes. Generated source must be reviewed, compiled, tested, and integrated in the destination repository. Framework and Selenium selections shape the output but do not replace engineering review.
What does a framework profile change?
It changes the generated test structure, imports and dependencies, setup and cleanup, assertion calls, and failure APIs for the selected C#, Java, or Python framework. Studio playback itself continues to use the built-in C# Selenium engine.
Where are the privacy, license, and third-party details?
The product website now includes the current Privacy Policy, Terms of Use, End-User License Agreement, and license, dependency, notice, and SBOM information. Pre-release legal text is clearly identified and should receive qualified review before public commercial distribution.
Evaluate the product
See how WebElemate turns a browser journey into automation your team can review, run, and own.
Start with the five-minute tutorial, then explore the controls that matter to your testing and engineering workflow.