MCP For Claude Marketplace
Brunel Engineering Prover

Brunel Engineering Prover MCP Connector for Claude

A+

A warehouse system hit 3x its tested throughput on a Tuesday morning. Sorting stations jammed in 12 minutes. 4 hours of standstill. $180K in delayed shipments. The operations manual said 'should handle expected growth.' Nobody tested what 'expected' meant — or when it stopped being true. Brunel specified Box Tunnel's gradient to exactly 1 in 100. Every brick course counted. This tool forces that rigor: analyze what breaks at 10x/100x, map integration interfaces, specify exact tolerances, quantify risks with probability and blast radius, and challenge precedent at your scale.

1 tools Official Updated Oct 1, 2026 Official Vinkius Partner
More Details Connect to Claude

AI agents design systems that work today and crumble tomorrow. They build components in isolation and call it 'decentralized.' They write 'should be efficient' and call it a specification. They say 'it might fail' and call it risk analysis.

The Problem

LLMs commit five engineering failures:

  • Scale Blind — 'The operation handles current volume.' Current volume is 200 orders per hour. The sales team just signed a partnership that triples volume next quarter. What breaks at 600/hour? Sorting stations? Verification clerks? Packaging capacity? Brunel designed the Great Western Railway for speeds no locomotive had yet achieved — he did not design for today's train.
  • Integration Neglected — 'Each department operates independently.' When the intake desk stops, what happens to sorting? Does sorting pause? Queue? Continue blindly? Overflow after 30 minutes and cascade to packaging? Brunel designed gauge, tunnels, stations, bridges, and rolling stock as ONE system. The broad gauge determined tunnel dimensions, platform heights, and loading gauge. Integration is not optional — it is the engineering.
  • Specification Absent — 'The process should be fast and reliable.' How fast? Measured how? At what consistency level? With how many concurrent orders? What happens when the specification is violated — escalate, slow down, or halt? Brunel specified Box Tunnel's gradient to 1 in 100 and counted every brick course. 'World-class operations' is not a specification.
  • Risk Unquantified — 'There might be issues with the supplier.' What is the probability? What data supports that? If they fail, how many orders are affected? How much revenue is lost? What is the specific mitigation and what residual risk remains? Brunel calculated exact yield loads on every girder of the Royal Albert Bridge. He did not say 'be careful.'
  • Precedent Worship — 'Nobody has structured a fulfillment center this way. The proven approach is a traditional sequential layout.' Nobody had built an iron ship before SS Great Britain. Nobody had built an underwater tunnel before the Thames Tunnel. The proven approach fails at your target volume — that is why you need a new one. What evidence shows the innovation works?

How It Works

5 Decision Pivots following Brunel's methodology:

  1. scaleAnalyzed — Current throughput measured. 10x bottleneck identified. 100x structural invalidation documented.
  2. integrationMapped — Component interfaces with contracts. Failure cascade documented. Backpressure paths mapped.
  3. specificationRigorous — Exact tolerances with measurement methods and violation consequences.
  4. riskQuantified — Probability × impact × mitigation for each specific failure scenario.
  5. precedentChallenged — Existing approach's scale failure identified. Innovation justified with evidence.

The Verdict Matrix

First Failing Pivot Verdict Meaning
scaleAnalyzed = false SCALE_BLIND Designed for current load only.
integrationMapped = false INTEGRATION_NEGLECTED Components built in isolation.
specificationRigorous = false SPECIFICATION_ABSENT Vague requirements, not tolerances.
riskQuantified = false RISK_UNQUANTIFIED 'Might fail' — no numbers.
precedentChallenged = false PRECEDENT_WORSHIP Refused to innovate at scale.
All pivots pass ENGINEERING_PROVEN Scaled. Integrated. Specified. Quantified. Innovated.
engineering-at-scalesystems-integrationspecification-rigorrisk-quantificationbrunelinfrastructure-architecturescale-analysis

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

validate_brunel_engineering

Think like Isambard Kingdom Brunel — who designed the Great Western Railway as ONE integrated system (gauge, tunnels, stations, bridges, locomotives) and built the SS Great Eastern at a scale no shipbuilder had attempted. You must: (1) ANALYZE SCALE — what breaks at 10x and 100x current demand? Not "handles today" — identify the FIRST bottleneck at each scale threshold. Which structural assumptions become invalid? Brunel designed for speeds no locomotive had achieved. Design for YOUR future scale, (2) MAP INTEGRATION — every component interface contract: what format, what timing, what quality standard, what happens on failure. Map failure cascades: if A fails, what happens to B, C, D? Map bottleneck propagation: where does overload accumulate? Brunel designed gauge, tunnels, stations, and bridges as ONE railway — not separate projects, (3) SPECIFY EXACTLY — precise tolerances, not "should be good." Delivery ≤ 48 hours (tolerance: 44-52h), error rate ≤ 0.5% (measured: weekly audit), cost ≤ $12/unit (consequence: contract breach). Brunel specified Box Tunnel gradient to 1 in 100. What are YOUR exact numbers? (4) QUANTIFY RISK — for each failure scenario: probability % (from data, not intuition) × impact (customers affected, revenue lost, reputation damage) × specific mitigation + residual risk. Brunel calculated exact yield loads on every bridge girder. "Might fail" is not quantified risk, (5) CHALLENGE PRECEDENT — when the existing approach fails at target scale, innovate with evidence. Name the existing method, show WHERE it breaks at your scale, propose the innovation, prove it works (pilot, calculation, comparable case), acknowledge the tradeoff. Brunel built the first iron ship when the world said it was impossible. If rejected, your engineering has a structural gap. Fix it before building.Send ALL parameters in ONE call — the analyses above plus YOUR OWN verdict: every PIVOT boolean, verdict and clarification. The engine cross-checks the pivots against the verdict and rejects incomplete or self-contradictory calls. Structured reflection tool for Brunel-grade systems engineering at scale. Forces the agent to analyze what breaks at 10x/100x demand, integrate components as one system with interface contracts and failure cascades, specify exact engineering tolerances with measurement methods, quantify risks with probability × impact × mitigation, and challenge precedent when existing approaches fail at target scale. Catches Scale Blindness (designing for today's load without 10x/100x analysis), Integration Neglect (components built in isolation without interface contracts or cascade mapping), Specification Absence ("should be good" instead of metric ≤ X, tolerance ± Y, measured how), Risk Handwaving ("might fail" instead of P(failure) × blast radius × mitigation), and Precedent Worship (following "the standard approach" when it structurally fails at target scale). Call once per large-scale system design, infrastructure plan, or organizational architecture

See how to talk to your AI agent using Brunel Engineering Prover.

Our warehouse processes current orders well — about 200 per hour with 95% completed within 4 hours. The operation should be reliable and scalable. We might have issues if volume grows, so we should be careful about capacity planning.

SCALE_BLIND — 'Handles current volume' is a snapshot, not scale analysis. 200 orders/hour at 4-hour completion — good. Now: at 2,000/hour, your sorting stations (6 staff, capacity ~50 orders/person/hour = 300/hour max) saturate at 300/hour. At 600/hour with your current layout, the intake area fills in 45 minutes with no outflow capacity. How many sorting stations do you have? What is each one's throughput? At what volume does the queue exceed your floor space? Brunel designed the GWR for speeds no locomotive had achieved. Design for YOUR 10x — what is the first station that jams at 2,000 orders/hour?

Scale: current 200/hour, 95% within 4h. At 10x: sorting stations (6 staff × 50/person = 300/hour max) saturate — time-studied March 18. At 100x: single-floor conveyor maxes at ~3,000/hour, multi-site needed. Integration: Intake → Verification (3 min) → Sorting (12 min) → Packaging (20 min). Sorting halts → intake pauses after 50-item buffer. Packaging halts → overflow staging, customer notified. Backpressure: 1 order per available sorting slot. Specification: 95% within 4h, 99% within 8h, ≥99.5% uptime. Measured by shift supervisor, 3 zones, hourly. Alert at >5h, overtime at >6h, halt intake at >8h. Risk: (1) Sorting overload — 15% probability Q3, 45min delay. Pre-hire 4 temp staff, 3% residual. (2) Supplier delay — 2%/month, 200 orders impacted. 2-week buffer + standby supplier. Precedent: linear scaling fails at 2K/hour (floor space). Innovation: zone-based parallel processing — pilot showed 40% throughput increase (50→70/person/hour). Tradeoff: 2 weeks cross-training per operator.

ENGINEERING_PROVEN — Brunel Method complete. Scale: 10x bottleneck identified (sorting at 300/hour, time-studied March 18), 100x invalidation documented (conveyor belt limits at 3,000/hour — multi-site required). Integration: 4-stage chain with contracts, failure isolation (intake pauses on sorting halt, overflow staging on packaging halt), and backpressure (1:1 intake-to-slot matching). Specification: exact tolerances (95% within 4h, 99% within 8h, 99.5% uptime), measurement method (supervisor logs, 3 zones, hourly), escalation chain (alert → overtime → halt). Risk: sorting overload quantified (15% Q3, 45min delay, pre-hire mitigation, 3% residual), supplier delay quantified (2%/month, buffer stock + secondary supplier, specialty materials gap acknowledged). Precedent: linear scaling failure proven at 2K/hour via blueprint, zone-based parallel processing tested in pilot (50→70 orders/person/hour), tradeoff acknowledged (cross-training cost, 2 weeks/operator). This is engineering.

Scale: at 10x, we add more sorting stations. At 100x, we need a bigger facility. Integration: each department is independently managed and self-contained — we can worry about handoffs later. Specification: the operation should have fast turnaround and high reliability.

INTEGRATION_NEGLECTED — Scale analysis passes: 10x add stations, 100x needs larger facility — trajectory identified. But integration fails catastrophically: 'independently managed and self-contained' and 'worry about handoffs later' is exactly the failure Brunel engineered against. The Great Western Railway worked because gauge, tunnels, stations, and rolling stock were designed as ONE system. When Sorting slows by 50% under holiday load, does Intake overflow in 30 minutes and cascade to Verification? When your packaging materials run out, does every upstream stage continue blindly filling a staging area with nowhere to go? Map your handoff contracts: item format, timing, capacity limits, overflow protocol, escalation threshold. NOW — not later.

No. Brunel engineered railways, ships, tunnels, and bridges — different domains, same method. This tool applies to any system that must survive its own success: warehouse operations, manufacturing lines, logistics networks, organizational processes, supply chains, service delivery systems. The 5 pivots — scale analysis, integration mapping, specification rigor, risk quantification, and precedent challenge — apply wherever a system must work at a scale it has not yet experienced.

Related Connectors

Enterprise Integration Complexity Scorer

Enterprise Integration Complexity Scorer MCP

4 tools Official

Quantify technical effort, risk, and resource needs for enterprise software integrations.

A+ View details →
Agent Workflow Bottleneck Analyzer

Agent Workflow Bottleneck Analyzer MCP

3 tools Official

Identifies performance bottlenecks and error risks in agentic pipelines.

A+ View details →
Copernicus Perspective Prover

Copernicus Perspective Prover MCP

1 tools Official

Your AI analyzed the problem from the default perspective and added workarounds when it did not fit. That is an epicycle — not a solution. Copernicus did not add more epicycles to Ptolemy's model. He moved the center from Earth to Sun. 40+ epicycles vanished. This tool forces default questioning, epicycle counting, alternative framing, observer shifting, and simplicity comparison.

A+ View details →
Delivery Integrity Prover

Delivery Integrity Prover MCP

1 tools Official

Forces AI agents to reflect on task execution, matching prompt requirements to actual changes, verifying logs, and declaring gaps before claiming completion.

A+ View details →