Use the Prompt
Add Your Brief
Add your context below. It will be included with the prompt when you copy or download.
Create an editable mobile-first app prototype using the existing project stack for [PROJECT] and [AUDIENCE]. The task is: Language Practice Session.
Required content and structure
Select a short practice session, answer a listening question with a text alternative, and review specific answers without inventing a proficiency score.
Screens and navigation
Build three connected views: Practice plan; Listening exercise; Session review. Keep the selected record visible in the screen title or summary. Support an explicit back action and preserve in-progress inputs when moving between views. Keep the primary action near the relevant content and account for mobile safe areas, a visible software keyboard and long translated labels. At desktop widths, adapt the composition without removing a step or changing the task.
Project inputs
Use [APPROVED_CONTENT] for names, records, labels, quantities and destinations. Ask for missing data instead of inventing operational claims. Demonstration records may use the fictional brand Lilt; label these as sample data. Explain which actions are local prototype interactions and which would require authentication, persistence, permissions or a backend. Do not connect external services or claim a transaction was completed unless that integration has been explicitly implemented and tested.
Design direction
Follow [BRAND_RULES]. When no brand is supplied, use lavender, white, deep plum, playful geometric marks. Make the workflow recognizable through its own content and controls, rather than repeating generic dashboard cards. Use a consistent spacing scale and clear text hierarchy. Include visible keyboard focus, accessible names, sufficient contrast and non-color status labels. Respect reduced motion and keep animated transitions optional.
Task-specific acceptance check
- Allow audio replay and skipped questions; distinguish an unanswered exercise from an incorrect response.
Interaction states
Implement the three-screen happy path with local sample data. Include a useful empty state, validation before completing an action, a loading or pending state and a recoverable failure state. Destructive changes require a review or undo path. Keep the state of a record consistent across all three views. Do not silently discard partially completed input or change a user's choice after a retry.
Delivery and review
Return the editable app views and explain the data model, any local storage behavior and unimplemented integrations. Test at 320px and a wider viewport, with keyboard navigation and long content. Use the task-specific acceptance check above to review the result before adding features. Keep demonstration behavior clearly distinct from a running production service.
Reference attachment and reconstruction goal
Attach reference.webp from this prompt’s download bundle alongside this prompt in OpenDesign. Build an editable result with a similar composition and hierarchy; do not embed the whole reference image as the finished interface. Preserve approved project content and brand rules if they conflict with fictional sample labels. If the image is missing or unreadable, ask for it instead of claiming visual fidelity.
Reference composition and content
Use the attached reference.webp as the visual reference for Language Practice Session. Before building, identify the major regions from top to bottom, their relative widths and heights, alignment, spacing, typography hierarchy, surface colors, borders, imagery and repeated components. Preserve those relationships rather than substituting a generic template. Required task content: Select a short practice session, answer a listening question with a text alternative, and review specific answers without inventing a proficiency score. Use the following task-specific layout guidance when translating the image into editable elements: Begin with Practice plan as the entry screen, give Listening exercise a clear primary action, and make Session review explain what happened and what can happen next. Use lavender, white, deep plum, playful geometric marks as a starting direction, adapting it to the supplied brand. Keep record identity and navigation context visible across the three screens. Prioritize thumb-accessible controls and readable labels over decorative panels; on wider screens, keep the workflow coherent rather than stretching a mobile composition edge to edge.
Visual direction from the example brief
lavender, white, deep plum, playful geometric marks.
Interaction specification
The reference is a static concept: the behavior below is a requirement to implement, not proof that the picture already works. Copy the customized prompt into your project and build the connected flow: Select a short practice session, answer a listening question with a text alternative, and review specific answers without inventing a proficiency score. Allow audio replay and skipped questions; distinguish an unanswered exercise from an incorrect response. For each visible control, specify its trigger, affected state and feedback; implement local changes with consistent state across related views. Give search and filters real results, selection controls an explicit active state, and save actions visible pending/success/failure feedback where relevant. Do not add unrelated workflows. Use proper links for navigation, buttons for actions and accessible names for icon controls. Do not imply a backend operation succeeded when only a local demo changed.
Visual comparison and delivery
- Compare the result with the attached image at matching proportions. Review major region placement, relative image sizes, line breaks, spacing, typography hierarchy, contrast, active states and content density before polishing decorative details.
- Then test the narrow layout without squeezing desktop text into tiny labels.
- Keep logos, photos and icons as separate assets and all meaningful labels editable. List any unavailable fonts/assets and any intentional departures. Return the working files plus a concise account of implemented behavior and remaining integrations; do not promise a pixel-identical result from the prompt alone.
Download Prompt + Image
Includes the complete prompt and full-size reference image. Attach both in OpenDesign to guide the layout.
How to Write a Language Practice Session Design Prompt
Use this language practice session design prompt to select a short practice session, answer a listening question with a text alternative, and review specific answers without inventing a proficiency score.
The reference is an AI-generated concept. The instructions below describe what to build and test; the image does not demonstrate working behavior.
Explain the Answer, Not an Invented Score
A language practice app should help someone understand an exercise result.
Skipped, unanswered and incorrect responses need different treatment, with accessible alternatives to audio. Begin with Practice plan as the entry screen, give Listening exercise a clear primary action, and make Session review explain what happened and what can happen next. Use lavender, white, deep plum, playful geometric marks as a starting direction, adapting it to the supplied brand. Keep record identity and navigation context visible across the three screens. Prioritize thumb-accessible controls and readable labels over decorative panels; on wider screens, keep the workflow coherent rather than stretching a mobile composition edge to edge.
From the Prompt
Build three connected views: Practice plan; Listening exercise; Session review.
Make It Your Own
Offer [SHORT_SESSION] with replayable audio and a text alternative. Review each response with the supplied answer explanation and avoid generating an unsupported proficiency rating.
Prepare the Language Practice Session Content
The exercise screen combines playback, speed, a question and selectable answers, with a typing alternative. Prepare the audio, transcript and answer model before building feedback.
Supply short exercises, answer explanations, audio files and matching text alternatives. Include unanswered and skipped responses so the review screen can distinguish them from incorrect answers without inventing a proficiency assessment.
From the Prompt
Select a short practice session, answer a listening question with a text alternative, and review specific answers without inventing a proficiency score.
Make It Your Own
For [PROJECT], use [APPROVED_CONTENT] as the source for this language practice session. Approved records and field labels for Practice plan; Listening exercise; Session review; the action and persistence rules; brand assets and permission requirements. Identify missing inputs before replacing them with demonstration content.
Skip a Listening Question
The review distinguishes correct responses, mistakes and unanswered items. Test that feedback refers to the actual response and that retry does not erase the previous learning record unintentionally.
The review should mark it as skipped rather than wrong. Audio failure should still leave a usable text-based route through the exercise. Test this specific boundary first: Allow audio replay and skipped questions; distinguish an unanswered exercise from an incorrect response. Then run the main flow with a long record title, no saved records and a failed action. A user should be able to tell what is saved, what is pending and what can be undone without relying on color. Verify that returning to the previous screen preserves entered data and that a screen reader encounters the title, status and primary action in a sensible order.
From the Prompt
Allow audio replay and skipped questions; distinguish an unanswered exercise from an incorrect response.
Make It Your Own
Review the language practice session against this requirement: Allow audio replay and skipped questions; distinguish an unanswered exercise from an incorrect response. Test this specific boundary first: Allow audio replay and skipped questions; distinguish an unanswered exercise from an incorrect response. Then run the main flow with a long record title, no saved records and a failed action. A user should be able to tell what is saved, what is pending and what can be undone without relying on color. Verify that returning to the previous screen preserves entered data and that a screen reader encounters the title, status and primary action in a sensible order. Preserve approved content and revise only the parts that fail this check.
Review Your Language Practice Session Result
The example is an AI-generated visual concept. Connected services, device permissions and persistent data require separate implementation and testing.
- Allow audio replay and skipped questions; distinguish an unanswered exercise from an incorrect response.
Follow-Up Prompt
Use this after the first result. It narrows the revision to the most important part of this task.
Refine only the Listening exercise to Session review transition. Allow audio replay and skipped questions; distinguish an unanswered exercise from an incorrect response. Show the relevant pending, failure and completed states with explicit labels. Preserve the approved navigation, typography and other screens. Return the changed views and describe how entered data survives a retry.
