Feynman Radical Simplification Prover

Feynman Radical Simplification Prover MCP Connector for Claude

A+

Stop your AI from hiding behind jargon — force it to explain simply, build from scratch, and justify every piece of complexity.

1 tools Official Updated Jun 28, 2026 Official Vinkius Partner

The Problem

Your LLM sounds smart. That is not the same as understanding.

Feynman reduced Quantum Electrodynamics — the most precise theory in physics — to three basic actions. He explained the Challenger disaster with a rubber O-ring in a glass of ice water. One gesture. One explanation. Devastating clarity.

LLMs do the opposite:

  • Use jargon to sound authoritative without explaining the mechanism
  • Hide behind "it's complicated" instead of finding the simple core
  • Recite best practices instead of building answers from scratch
  • Express high confidence without identifying where they might be wrong
  • Add complexity that carries no explanatory power

How It Works

1 tool: validate_radical_simplification

The agent must fill 5 reflection fields and commit to 5 boolean Decision Pivots:

Pivot Question
jargonEliminated Can you explain this WITHOUT domain terminology?
simplificationAchieved Can a NON-EXPERT follow your explanation?
constructionDemonstrated Can you BUILD the answer from scratch?
selfDeceptionExposed Have you identified where you might be FOOLING YOURSELF?
complexityJustified Does every piece of complexity ADD explanatory power?

Verdict Matrix

Pivot Failure Verdict
Pivot 1 fails JARGON_HIDING
Pivot 2 fails SIMPLIFICATION_FAILED
Pivot 3 fails CONSTRUCTION_ABSENT
Pivot 4 fails SELF_DECEPTION
Pivot 5 fails COMPLEXITY_UNJUSTIFIED
All pass UNDERSTANDING_PROVEN

Why It Works

Tool calls are obligations — instructions are suggestions. The agent cannot skip the simplification check. ~70 semantic traps catch jargon-hiding ("synergize", "paradigm shift"), complexity-refuge ("it's too complex to simplify"), recitation ("best practices dictate"), self-deception ("I am confident that"), and gratuitous complexity ("myriad considerations").

Feynman: "The first principle is that you must not fool yourself — and you are the easiest person to fool."

structured-reasoningdecision-pivotssimplificationfeynman-techniqueunderstandinganti-jargon

1 tools expose this connector's capabilities to your AI agent.

validate_radical_simplification

You must: (1) ELIMINATE JARGON — rewrite the explanation using only words a bright teenager understands. Feynman: "I learned very early the difference between knowing the name of something and knowing something." If you cannot explain it without the technical term, you do not understand it, (2) SIMPLIFY TO CORE — reduce to the ONE mechanism that makes this work. Feynman: QED = 3 actions (a photon goes from place to place, an electron goes from place to place, an electron emits or absorbs a photon). What are YOUR 3 actions? If "it is too complex to simplify," you do not understand it, (3) CONSTRUCT FROM SCRATCH — build the answer step by step from bedrock certainties. Feynman: "What I cannot create, I do not understand." Do not recite — derive. Start from what you know for certain and build to the conclusion, (4) EXPOSE SELF-DECEPTION — identify where you might be fooling yourself. Feynman: "The first principle is that you must not fool yourself — and you are the easiest person to fool." Name your weakest point, your most convenient belief, your blind spot, (5) JUSTIFY COMPLEXITY — prove every layer of complexity earns its place. Feynman on Challenger: O-ring in ice water. One gesture, entire disaster explained. What is YOUR O-ring? Which layers can be removed without losing understanding? If rejected, fix the specific understanding gap the engine identifies. Structured reflection tool that forces Feynman-level radical simplification. The agent must eliminate jargon (knowing the name is not knowing the thing), simplify to the irreducible core mechanism, construct answers from scratch instead of reciting them, expose self-deception ("you are the easiest person to fool"), and justify every layer of complexity (the O-ring test — one gesture should explain the failure). Catches Jargon Hiding (technical terms masking lack of understanding), Simplification Failure (inability to reduce to first principles), Construction Absence (reciting instead of deriving), Self-Deception (convenient beliefs masking weak reasoning), and Unjustified Complexity (layers that add no explanatory power). Call once per explanation, analysis, or recommendation

See how to talk to your AI agent using Feynman Radical Simplification Prover.

We need a microservices architecture with event-driven choreography and CQRS to handle our scaling challenges.

JARGON_HIDING — Three technical terms in one sentence. What does 'event-driven choreography' mean in plain language? If the explanation collapses without the jargon, the understanding was never there.

I am certain our migration will succeed. The plan covers every scenario.

SELF_DECEPTION — 'I am certain' and 'every scenario' in the same breath. Feynman: the first principle is that you must not fool yourself. Where SPECIFICALLY could this plan fail?

Our performance issue is too complex to explain simply. There are too many interacting variables.

SIMPLIFICATION_FAILED — 'Too complex to explain simply' means you do not understand it yet. Feynman simplified quantum physics. What is the ONE bottleneck? Find your O-ring.

No. It forces the agent to produce its OWN simplified explanation and then validates consistency. If the agent claims jargon is eliminated but uses 'synergize' or 'paradigm shift,' the engine rejects. The simplification must be genuine — not jargon renamed.

Related Connectors