expired-patent-scout
Expired Patent Scout
Find useful technical ideas whose relevant patent term appears to have ended in the user's target markets, then explain what product opportunities merit further work. This is a research shortlist, not legal clearance to manufacture or sell. Recommend patent counsel review of selected implementations and jurisdictions before commercial reliance.
Start with a short intake
Ask about the user's interests before choosing the search targets. Reuse answers already present. Use an available question tool or concise chat questions, without requiring a particular vendor's tool name. Group the essential questions:
- What should we investigate? Brands/companies, specific products or model names, technology areas, or problems the user wants to solve. Product/brand links and known patent numbers are helpful but optional. A category is enough; named brands are not required.
- Where would it be made and sold? Countries for manufacture, import, sale, and use. Ask for the initial research market if the user is undecided; do not label something globally expired based on one country's record.
- What counts as recent? Offer the last 12 months, 24 months, or a custom range. Use the last 24 calendar months as the stated working default when this preference is omitted, and record absolute dates plus an as-of date using the user's current date. Ask whether a separate upcoming-expiration list would help; include it only when requested.
Ask about constraints only when they affect ranking: intended customer/use case, price range, manufacturing capabilities, appetite for electronics/software, utility mechanisms versus product appearance, and research depth. Default to a focused shortlist of up to 10 strong opportunities, with fewer when the evidence supports fewer. Do not fill a quota with weak or unrelated patents.
If the target area is missing, keep intake pending and do not invent brands. If markets are missing, continue technical/brand discovery but leave jurisdiction-specific conclusions unresolved. Do not ask for brands now when the task is only to create or edit this skill.
Save the agreed scope in research.json: targets, countries/activities, date window, patent types, exclusions, optional future window, constraints, and as-of date. State material assumptions before searching.