ux-ia-review

name	ux-ia-review
description	Review a provided UI screenshot for usability and information architecture quality, then propose concrete structural changes with rationale and generate a copy/paste prompt for a coding agent. Use when the user asks for UX review, IA review, layout structure critique, navigation clarity feedback, or task-flow improvement based on a screenshot or mockup.

UX IA Screenshot Review

Review one or more screenshots to diagnose usability and information architecture issues, then convert findings into actionable structural changes and a coding-agent-ready prompt.

Scope Rules

• Focus on usability and information architecture only.
• Analyze hierarchy, grouping, labeling, navigation, task flow, discoverability, and cognitive load.
• Avoid visual-style critique (colors, typography, icon style, branding, aesthetics) unless it directly blocks usability.
• Avoid technical implementation instructions (frameworks, CSS, component APIs, code-level detail).

Required Inputs

• Use at least one screenshot provided by the user.
• Use optional context when available: user type, primary task, device type, and product goal.
• State explicit assumptions when context is missing instead of blocking progress.

Workflow

1. Map the Current Structure

• Identify major regions: global navigation, local navigation, page header, content areas, sidebars, utilities, and calls to action.
• Describe the apparent primary task and secondary tasks supported by the screen.
• Note the reading/order path implied by layout and emphasis.

2. Evaluate Information Architecture

• Check grouping and chunking of related information.
• Check labeling clarity and whether terms match user intent.
• Check findability of key actions and key information.
• Check hierarchy depth and whether users can orient themselves quickly.

3. Evaluate Usability

• Check whether the next action is obvious at each stage.
• Check friction points, ambiguity, duplicated decisions, and dead ends.
• Check whether the interface likely increases cognitive load unnecessarily.
• Check whether navigation and content structure support the main workflow.

4. Define Change Proposals

For every meaningful issue, provide:

• What exists now (current structure/behavior).
• What should change (target structure/behavior).
• Why the change improves usability or IA.
• Priority (High, Medium, Low) based on impact on key task completion.

5. Generate Coding-Agent Prompt

Produce a final prompt that tells a specialized coding agent what to change, not how to code it.

• Include explicit goals and constraints.
• List required UX/IA changes in priority order.
• Include rationale to preserve intent.
• Include acceptance criteria phrased in UX/IA outcomes.
• Exclude technical implementation details.

Output Format

Always return sections in this order:

1) What I Found

• Summarize the current screen structure in 4-8 bullets.
• State the primary task the user appears to be trying to complete.
• State assumptions if context is missing.

2) What Needs To Change

Use a table with these columns:

• Issue
• Current State
• Proposed Change
• Why This Improves UX/IA
• Priority

3) Structural Direction (Optional)

• Provide a compact proposed IA outline when re-organization is substantial.
• Use simple bullets or numbered lists for navigation/content structure.

4) Copy/Paste Prompt For Coding Agent

Return a fenced text block with this shape:

You are improving the UX and information architecture of an existing interface.

Objective:
- [Outcome-focused objective]

Constraints:
- Focus on usability and information architecture only.
- Do not spend effort on visual styling changes unless required for usability.
- Implement WHAT should change; choose HOW to implement.

Required changes (in priority order):
1. [Change request]
2. [Change request]
3. [Change request]

Rationale to preserve:
- [Key UX/IA reasoning]

Acceptance criteria:
- [Observable UX/IA outcome]
- [Observable UX/IA outcome]
- [Observable UX/IA outcome]

Quality Bar

• Tie every recommendation to evidence visible in the screenshot.
• Translate diagnosis into concrete structural change requests.
• Keep findings specific and actionable; avoid generic UX advice.
• Keep language implementation-agnostic for the final coding-agent prompt.