Breakthrough Ideation Prover MCP Connector for Claude

AI agents have a fundamental ideation problem. When asked for ideas, they default to the safest, most obvious answer — the first thing a Google search would return. When pushed to be creative, they swing to the opposite extreme: hallucinating impossible solutions built on nonexistent technology, ignoring every real-world constraint, and presenting grand visions with no path to execution.

The Problem It Solves

AI ideation fails on five axes:

• Conventional thinking — "Use social media marketing." "Hire more engineers." "Implement a CRM." Ideas anyone would suggest. Zero departure from the obvious.
• Fantasy ideation — "Use quantum computing to solve this." "Build an AI that reads minds." Creative-sounding but physically, economically, or technically impossible with current technology.
• Constraint blindness — A brilliant idea that ignores the $10K budget, the 2-person team, the 3-month deadline, or the regulatory environment. Innovation in a vacuum.
• Vision without path — "Disrupt the industry with a paradigm-shifting platform." HOW? What's step 1? What happens tomorrow morning? Who does what?
• Unrealizable commitment — The idea hits a fatal blocker (regulatory, physical, economic) and the agent either ignores it or gives up. Instead of finding a way around it.

How It Works

Breakthrough Ideation Prover uses 5 Decision Pivots — boolean checkpoints that force the agent to prove its idea is BOTH radical AND achievable:

1. conventionChallenged — Does this idea genuinely depart from the obvious? You must NAME the conventional approach to prove you've transcended it.
2. factuallyGrounded — Is it built on real facts? Cite the company, the research, the case study, the technology benchmark. Not "studies show" — WHICH studies?
3. constraintsIntegrated — Are budget, timeline, team, regulation, and technical maturity acknowledged AND woven into the idea? Not ignored.
4. implementationPathMapped — Is there a concrete roadmap? Week 1 action, Month 1 milestone, Quarter 1 validation point. WHO does WHAT by WHEN?
5. feasibilityProven — For EACH critical blocker, is there a SPECIFIC workaround? Not "easily overcome" — name the blocker, name the mitigation.

Why It Works

• The convention anchor. By forcing the agent to explicitly name the conventional approach FIRST, it creates a cognitive anchor that the breakthrough must demonstrably surpass. You can't claim innovation without naming what you're innovating FROM.
• The grounding requirement. Radical ideas grounded in overlooked facts are the most powerful kind. Tesla didn't invent the electric car — they grounded it in lithium-ion battery cost curves that others overlooked. The factual basis requirement ensures ideas are built on reality, not imagination.
• The feasibility obligation. Most "creative" AI output dies at implementation. This tool demands that for EVERY blocker, the agent proposes a SPECIFIC workaround. Not hand-waving — concrete mitigations. If the idea is truly worthwhile, there IS a way.
• Semantic traps. The engine catches ideation-specific anti-patterns: strawman conventional approaches ("do nothing"), vague factual claims ("studies show"), constraint dismissals ("no real barriers"), abstract roadmaps ("Phase 1: Plan"), hand-wavy feasibility ("easily overcome"), and non-committal conclusions ("worth exploring").