Deep Analyst Prover MCP Connector for Claude
A+AI gives surface analysis — restates the question, misses hidden assumptions, uses single-lens thinking. This tool forces multi-model depth: First Principles decomposition, Second-Order cascades (3 levels), Steelmanning (Ideological Turing Test), Inversion, and Premortem risk mapping....
AI models give surface-level analysis. They restate the question instead of decomposing it, miss hidden assumptions, use single-lens thinking, fail to stress-test from the opponent's view, see only first-order effects, and produce generic conclusions that could apply to any problem.
The Problem
AI analysis fails on seven axes:
- Surface analysis — restates "how should we price?" as "pricing is important" without decomposing into sub-problems.
- Hidden assumptions — doesn't surface the beliefs that, if wrong, invalidate the entire analysis.
- Single-model thinking — applies one lens when 3+ mental models would reveal completely different insights.
- No opposition — doesn't steelman the opposing view. Strawman ("some might say") is not steelman.
- Cascade blindness — sees immediate effects but misses Level 2 and Level 3 consequences.
- No risk mapping — skips premortem analysis that makes risks 30% more identifiable (Klein, 2007).
- Generic conclusion — summary instead of synthesis. Repeats what was said instead of combining models into a novel insight.
How It Works
6 Decision Pivots force genuine intellectual depth:
- decomposed — Problem broken into 3-5 atomic sub-problems? Not just restated?
- assumptionsExplicit — 3-5 load-bearing assumptions listed with "if wrong" consequences?
- multiModelApplied — 3+ mental models stacked? First Principles + Second-Order + Inversion minimum?
- oppositionSteelmanned — STRONGEST case against your conclusion? Ideological Turing Test?
- cascadesIdentified — "And then what?" 3 levels deep? L1 → L2 → L3?
- risksPremortem — "It failed spectacularly. Why?" 3+ specific failure paths?
Related Connectors
Nikola Tesla Inventor Prover MCP
An AI suggested 'build a pilot and iterate' for a complex system requiring 18 months of structural validation. This tool forces it to simulate the complete system in its mind, prove it mathematically, and find resonance instead of brute-force scaling.
YC Startup Prover MCP
Most startups die building solutions nobody asked for. A founder pitched for 20 minutes about their proprietary algorithm — never mentioned a single user, a single pain point, a single dollar of willingness to pay. That pitch would be rejected in 30 seconds at Y Combinator. This tool forces five PG-level axes: problem discovery, unscalable beginnings, metric discipline, user obsession, and core value focus.
Context Integrity Prover MCP
AI forgets the original goal and hallucinates scope. This engine is a 6-pivot trap that forces the LLM to prove it hasn't drifted from the user's explicit constraints.
Product Discovery Prover MCP
Block engineering waste. This gatekeeper demands hard data, behavioral segments, and proven willingness-to-pay before a single line of code is written.