What an option looks like
Input
The chart’s closest limitation: dynamic adjustment of the rate limit upon burst detection
Output
| Token bucket, fixed refill · drains per request, refills on a timer, so the limit never adjusts to traffic shape | absent verified |
| Static tiered caps · three fixed rate tiers chosen at deploy time, no runtime adjustment at all | absent verified |
| Client-side shedding · clients self-throttle from response headers, server never detects a burst | different approach |
Illustrative example, software-scoped. Brainstorming works on any domain, but the casebook precedents and deepest vocabulary are software. It is an Engineer and Consultant feature: 20 runs/mo on Engineer, 100/mo priority on Consultant.
Start with a check See plans
How the brainstormer works
The brainstormer runs two generation passes and a verification gauntlet, in this order.
Pass one: proven patterns
The generator works from the clearance chart of your check and retrieves precedents from the casebook: design-around alternatives that past users proposed, that the scoring engine verified as real escapes, and that came from public-patent checks. The result is alternatives anchored in what has actually worked before, not generic suggestions.
Pass two: fresh ideas
A second pass generates deliberately divergent options with the casebook anchors removed, so the two passes do not converge on the same idea. If every verified escape in history pushed processing to the edge, the fresh pass is the one allowed to ask what happens if you centralize instead.
The gauntlet: every option gets re-scored
Each alternative declares the exact claim limitation it claims to break, quoted verbatim from the patent. The scoring engine then checks that claim: does this alternative actually read as absent or different-approach against that limitation? Options that score too similar are regenerated once, and anything that still fails is shown flagged, never silently dropped. You also get a prosecutor attack for each option: the strongest argument that it still infringes. Reading that argument before your attorney does is the cheapest legal review you will ever get.
The caution the engine cannot do for you
A design-around is only real if it breaks at least one limitation of every claim that matters, and a near-equivalent substitution can land you back inside the claim under the doctrine of equivalents even after a clean chart. The engine scores the mechanics; the judgment call is yours and your attorney’s.
What happens to my claims and design?
Privacy is the same as the check: your claims and design go to the scoring engine and back to you. Casebook capture only happens when the claims came straight from a public patent lookup, never from pasted text.
Questions about brainstorming
- Why do I need a check first?
- The brainstormer reads the clearance chart to learn which limitations your design is closest to. Without the chart it would be guessing what to escape.
- What are proven patterns vs fresh ideas?
- Proven patterns are grounded in verified design-arounds from past public-patent runs. Fresh ideas are deliberately divergent, with no anchors. You get both in one run.
- Can an alternative be wrong?
- Yes, and the engine says so. Each option is re-scored against the exact limitation it claims to break; failures are flagged with their scores, not hidden.
- Which plans include it?
- Engineer includes 20 brainstorm runs a month, Consultant 100 with priority. Free accounts do not include brainstorming.