information-architecture-review

name	Information Architecture Review
description	Child skill under Build Quality UI/UX for the structure-first lens: map the user's goal, the system's conceptual model, the primary flows, navigation and wayfinding, labels and grouping, feedback, and error recovery — before visual polish. Returns a structural map plus evidence-backed findings.
skill_type	strategy
altitude	domain
activation	progressive
parent_id	build_quality_ui_ux
depth	1
preserve_markdown	true
legacy_paths	["product-and-design.information-architecture-and-interaction-fundamentals.skill","product-and-design/information-architecture-and-interaction-fundamentals"]
reference_modules	[{"id":"information_architecture_review.ia_heuristics","name":"IA & Interaction Heuristics","summary":"The deep canon — Norman's gulfs and seven stages of action, affordance vs signifier, Cooper's conceptual/mental/implementation model gap, the IA-relevant Nielsen heuristics, Morville's findability facets, the wayfinding triad, and label/grouping and recovery checklists.","when_to_load":["Before producing structural findings, to apply named heuristics rather than generic UX intuition.","When the user disputes a finding and you need the underlying principle and source."],"path":"references/ia-heuristics.md"}]
path	apps/web/src/lib/services/agentic-chat/tools/skills/definitions/information_architecture_review/SKILL.md

Information Architecture Review

Identity

The structure-first lens. Before anyone asks whether a screen is pretty, this skill asks whether the user can understand it: do they know where they are, what things mean, what is actionable, what just happened, and how to get back. Run this lens first when a feature, flow, or workspace feels confusing — visual polish on a broken conceptual model just makes the confusion look nicer.

This is a strategy skill at domain altitude. This is the holistic pass: build a map of the system's structure and the user's path through it, then report structural fixes before cosmetic ones. The deep heuristics live in the information_architecture_review.ia_heuristics reference module — load it before producing findings.

Activation

• Reviewing a feature, flow, navigation model, dashboard, workspace, settings area, builder, or multi-step task that feels confusing, even when the individual screens look acceptable.
• The user asks about information architecture, labels, affordances, signifiers, conceptual models, navigation, wayfinding, feedback, or error recovery.
• A screen looks fine but users may not know what to do, where they are, what changed, or how to recover.
• ui_ux_quality_review (or the parent router) escalated a "conceptual confusion" finding here.
• You need to understand the whole structure before a redesign, not just audit one screen's pixels.

Judgment

The decision spine. When the pass surfaces something, this is what you reason with.

The named lenses

Apply these named lenses rather than generic intuition (deep canon in information_architecture_review.ia_heuristics):

• Gulfs (Norman): is the gulf of execution bridged (can the user see how to act?) and the gulf of evaluation bridged (can the user tell what the system did?).
• Model gap (Cooper): is the interface presenting a clean conceptual model, or is it leaking the implementation model (database tables, internal states, system jargon)?
• Labels & grouping: do labels use the user's words, carry information scent, and group by user task rather than internal structure?
• Affordances & signifiers: does what looks actionable match what is actionable, and does each action signify its result?
• Wayfinding triad (Morville): can the user always answer Where am I? Where can I go? How do I get back?