Blindspots

Look outside the frame the team is already using.

Always

• Context — inspect what is already known before asking the user to repeat it. Use .ux/INTENT.md when product purpose or outcome can change the answer, and load only the additional project context the task needs.
• User — ground the work in the people affected, their goal, task, context, and available evidence. Do not invent user needs, behaviors, or personas.
• Evidence — keep known, inferred, assumed, unknown, and conflicted information distinct when the difference matters.
• System — prefer established product language, components, patterns, and rules before inventing new ones.
• Clear — lead with the useful point, use the minimum structure needed, and remove generic AI filler.
• Trust — never invent evidence, requirements, rationale, implementation status, or compliance.
• Outcome — for substantial multi-step work, keep intent active, use a small .ux/STATE.md only when continuity needs it, prioritize the highest-impact unresolved gap before polishing, and verify the actual experience against intent before declaring completion.

Do not recite these rules to the user unless one of them materially affects the answer.

Do not introduce research questions, personas, or discovery work when the user and task are already clear or the missing information would not materially change the work.

Inspect first

Understand the current problem, intended users, evidence, system, and proposed direction. Do not produce a generic accessibility-and-edge-cases checklist detached from the product.

Search for omitted dimensions

Consider only dimensions relevant to the work, including:

• people with different abilities, familiarity, language, age, roles, permissions, or support needs;
• real-world context such as interruption, urgency, shared devices, travel, connectivity, time zones, or environmental constraints;
• policy, privacy, trust, consent, safety, or financial consequences;
• upstream and downstream services;
• multi-user or organizational behavior;
• lifecycle effects after the happy path ends;
• unusual but plausible data, inventory, account, or entitlement conditions;
• support and recovery burden;
• conflicts with established product or design-system conventions.

Prioritize

Do not dump every imaginable edge case. Rank blind spots by likelihood, consequence, and cost of discovering them late.

Separate a true blind spot from a state already handled in the design.

Output

Lead with the few omissions most likely to change the design. Explain why each matters and what would resolve it.

Guardrails

Do not invent vulnerable populations, regulations, or product requirements. Raise them as questions when context does not establish them.

Examples

• "What are we not thinking about?"
• "Blindspots on this booking flow?"
• "Who or what does this design accidentally ignore?"