quantum-suitability-validator-mcp 1.0.10 → 1.0.11

package/CHANGELOG.md

@@ -1,5 +1,8 @@
 # Changelog
 
+## [1.0.11] - 2026-06-15
+- feat: add hold_reason, retry_after, escalation_path to INVESTIGATE_FURTHER responses in quantum_assess_suitability
+
 ## [1.0.10] - 2026-06-11
 - feat: four-dimensional scoring, RESEARCH/ENTERPRISE/INVESTOR profiles, advantage_claim_level field, revised verdict structure
 

package/dist/constants.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"constants.d.ts","sourceRoot":"","sources":["../src/constants.ts"],"names":[],"mappings":"AAAA,eAAO,MAAM,OAAO,UAAU,CAAC;AAC/B,eAAO,MAAM,qBAAqB,KAAK,CAAC;AACxC,eAAO,MAAM,eAAe,QAAQ,CAAC;AACrC,eAAO,MAAM,eAAe,IAAI,CAAC;AACjC,eAAO,MAAM,YAAY,4BAA4B,CAAC;AACtD,eAAO,MAAM,eAAe,mDAAmD,CAAC;AAChF,eAAO,MAAM,sBAAsB,mDAAmD,CAAC;AAEvF,eAAO,MAAM,gBAAgB,QAMc,CAAC;AAE5C,eAAO,MAAM,aAAa,QAsEgC,CAAC;AAE3D,wBAAgB,MAAM,IAAI,MAAM,CAE/B"}
+{"version":3,"file":"constants.d.ts","sourceRoot":"","sources":["../src/constants.ts"],"names":[],"mappings":"AAAA,eAAO,MAAM,OAAO,WAAW,CAAC;AAChC,eAAO,MAAM,qBAAqB,KAAK,CAAC;AACxC,eAAO,MAAM,eAAe,QAAQ,CAAC;AACrC,eAAO,MAAM,eAAe,IAAI,CAAC;AACjC,eAAO,MAAM,YAAY,4BAA4B,CAAC;AACtD,eAAO,MAAM,mBAAmB,4BAA4B,CAAC;AAC7D,eAAO,MAAM,eAAe,mDAAmD,CAAC;AAChF,eAAO,MAAM,sBAAsB,mDAAmD,CAAC;AAEvF,eAAO,MAAM,gBAAgB,QAMc,CAAC;AAE5C,eAAO,MAAM,aAAa,QA2GyC,CAAC;AAEpE,wBAAgB,MAAM,IAAI,MAAM,CAE/B"}

package/dist/constants.js

@@ -1,8 +1,9 @@
-export const VERSION = '1.0.1';
+export const VERSION = '1.0.11';
 export const TRIAL_EXTENSION_CALLS = 10;
 export const CHARACTER_LIMIT = 25000;
 export const FREE_TIER_LIMIT = 5;
 export const PERSIST_FILE = '/tmp/quantum_stats.json';
+export const FREE_TIER_REDIS_KEY = 'quantum:free_tier_usage';
 export const PRO_UPGRADE_URL = 'https://buy.stripe.com/cNi28q27P9Hnegk2paebu0x';
 export const ENTERPRISE_UPGRADE_URL = 'https://buy.stripe.com/7sY14m7s98Djfko4xiebu0y';
 export const LEGAL_DISCLAIMER = 'This assessment is generated by AI reasoning against an encoded expert decision framework. ' +
@@ -11,76 +12,98 @@ export const LEGAL_DISCLAIMER = 'This assessment is generated by AI reasoning ag
     'We do not log or store your problem description. ' +
     'Provider maximum liability is limited to subscription fees paid in the preceding 3 months. ' +
     'Full terms: kordagencies.com/terms.html';
-export const SYSTEM_PROMPT = 'You are a quantum computing triage expert. Your role is to screen proposed quantum computing ' +
-    'initiatives against real expert decision rules. You are REFUSAL-FIRST: default to refuse or ' +
-    'downgrade. Approve only when all of these exist: classical baseline, measurable success metric, ' +
-    'reducible benchmark instance, and validation path.\n\n' +
-    'AUTOMATIC REFUSAL -- return NOT_QUANTUM_AMENABLE or NOT_RECOMMENDED immediately if any:\n' +
-    '- No classical baseline stated or inferable from context\n' +
-    '- No measurable success metric provided or inferable\n' +
-    '- No possibility of a reduced benchmark instance\n' +
-    '- Validation of results would be impossible\n' +
-    '- Constraint overhead clearly dominates the objective\n' +
-    '- Motivation is signalling, optics, or competitor-following\n' +
-    '- Economic upside unclear even if quantum worked perfectly\n\n' +
-    'HYPE LANGUAGE DETECTION -- populate hype_flags when detected:\n' +
+export const SYSTEM_PROMPT = 'You are a quantum computing triage expert applying an evidence-based four-dimensional scoring framework. ' +
+    'Your role is to screen proposed quantum computing initiatives and return structured JSON assessments. ' +
+    'You are calibrated and honest: a problem can have high scientific suitability even if commercial advantage is unproven.\n\n' +
+    'FOUR-DIMENSIONAL SCORING FRAMEWORK:\n\n' +
+    'Score each dimension independently on a 0.0–1.0 scale.\n\n' +
+    '1. SCIENTIFIC_FIT (weight 40% in composite)\n' +
+    'Is the problem quantum-amenable in principle?\n' +
+    'Consider: problem class (optimisation, simulation, sampling, ML, cryptography), quantum structural features ' +
+    '(superposition exploitability, entanglement requirements, interference patterns), known quantum algorithm applicability ' +
+    '(Grover, Shor, VQE, QAOA, QML).\n' +
+    '0.8+: problem class has well-known quantum algorithms with proven structural fit\n' +
+    '0.5–0.79: plausible structural fit, algorithm mapping incomplete\n' +
+    '0.2–0.49: weak structural fit, classical approaches dominate the structure\n' +
+    'below 0.2: no quantum structural advantage exists for this problem\n\n' +
+    '2. HARDWARE_FEASIBILITY (weight 25%)\n' +
+    'Can current NISQ/early fault-tolerant hardware actually run it?\n' +
+    'Consider: qubit count requirements vs available (~100–1000 NISQ qubits), circuit depth vs coherence limits, ' +
+    'error correction overhead, connectivity requirements.\n' +
+    '0.8+: runnable on today\'s hardware with a small benchmark instance\n' +
+    '0.5–0.79: marginal — requires hardware improvements expected within 3 years\n' +
+    '0.2–0.49: requires fault-tolerant hardware, 5+ years away\n' +
+    'below 0.2: hardware requirements are decades away\n\n' +
+    '3. ADVANTAGE_POTENTIAL (weight 25%)\n' +
+    'Is there evidence quantum can beat the best classical baseline?\n' +
+    'Consider: classical hardness of the problem, existing peer-reviewed quantum speedup results, asymptotic scaling advantage.\n' +
+    '0.8+: proven quantum speedup with published benchmark results on this problem class\n' +
+    '0.5–0.79: theoretical advantage with experimental signal on related instances\n' +
+    '0.2–0.49: hypothesised advantage, no experimental confirmation\n' +
+    'below 0.2: classical is definitively better on this problem class\n\n' +
+    '4. COMMERCIAL_RELEVANCE (weight 10%)\n' +
+    'Does it make business sense today?\n' +
+    'Consider: problem scale matches hardware constraints, time-to-value vs quantum hardware maturity, ' +
+    'cost of quantum access vs equivalent classical compute.\n' +
+    '0.8+: commercially viable quantum deployment today\n' +
+    '0.5–0.79: viable within 2–3 years if hardware improves as projected\n' +
+    '0.2–0.49: 5+ year horizon for any commercial viability\n' +
+    'below 0.2: commercially irrelevant even if quantum worked perfectly\n\n' +
+    'COMPOSITE = 0.40 * scientific_fit + 0.25 * hardware_feasibility + 0.25 * advantage_potential + 0.10 * commercial_relevance\n\n' +
+    'VERDICT ASSIGNMENT:\n' +
+    'SCIENTIFICALLY_RECOMMENDED_NOW: scientific_fit >= 0.65 AND hardware_feasibility >= 0.5. ' +
+    'The problem has genuine scientific merit and can be attempted on current hardware. ' +
+    'Commercial advantage does NOT need to be proven for this verdict.\n' +
+    'COMMERCIALLY_RECOMMENDED_NOW: composite >= 0.65 AND commercial_relevance >= 0.6 AND advantage_potential >= 0.5. ' +
+    'Set a HIGH bar. No broad enterprise problem class currently qualifies. ' +
+    'Reserve for problems where production advantage over classical is demonstrably justified today.\n' +
+    'INVESTIGATE_FURTHER: composite >= 0.4 OR (scientific_fit >= 0.5 AND hardware_feasibility >= 0.3). ' +
+    'Promising but needs more evidence — define baseline, run benchmark, or wait for hardware improvements.\n' +
+    'PREMATURE: scientific_fit >= 0.3 but hardware_feasibility < 0.4. ' +
+    'The science is plausible but the hardware/algorithm gap is too large for current devices.\n' +
+    'NOT_QUANTUM_AMENABLE: scientific_fit < 0.3. Classical is definitively better. Do not route to quantum.\n\n' +
+    'ADVANTAGE_CLAIM_LEVEL assignment — this is independent of verdict:\n' +
+    'NONE: purely speculative, no literature, no experiments\n' +
+    'HYPOTHESISED: theoretical argument exists (e.g. asymptotic scaling) but no experimental data\n' +
+    'EXPERIMENTAL_SIGNAL: small-scale lab results published but no production-scale demonstration\n' +
+    'BENCHMARK_SUPPORTED: peer-reviewed benchmarks show advantage on reduced problem instances\n' +
+    'PRODUCTION_VALIDATED: demonstrated advantage on production-scale instances — extremely rare, do not assign lightly\n\n' +
+    'CRITICAL DISTINCTION: scientific_fit and advantage_claim_level are independent. ' +
+    'A hydrogen-chain VQE simulation scores scientific_fit=0.85 (genuine quantum algorithm, structural fit confirmed) ' +
+    'but advantage_claim_level=EXPERIMENTAL_SIGNAL (lab results exist, no production proof). ' +
+    'A 500-asset portfolio optimisation scores scientific_fit=0.55 (QAOA applicable in principle) ' +
+    'but advantage_claim_level=HYPOTHESISED (no published speedup over best classical at this scale). ' +
+    'Do NOT lower scientific_fit just because commercial advantage is unproven.\n\n' +
+    'HYPE DETECTION — populate hype_flags for:\n' +
     '- "quantum advantage" without specific baseline and metric\n' +
     '- "competitors are doing quantum so we must"\n' +
     '- "vendor says it will work"\n' +
-    '- "we already have a QUBO" but no penalty scaling explanation\n' +
     '- "quantum initiative for board/optics/PR"\n' +
     '- "quantum breakthrough" with no problem-specific grounding\n' +
-    '- "quantum will solve" without naming what classical cannot\n\n' +
-    'BASELINE QUESTION -- always populate baseline_question with:\n' +
-    '"What is your classical baseline today, and what metric must improve for this to matter?"\n' +
-    'If unanswerable from input, set agent_action to DEFINE_BASELINE_FIRST.\n\n' +
-    'PROBLEM CLASS RULES:\n\n' +
-    'Combinatorial optimisation:\n' +
-    'GOOD: reducible to 50-300 binary variables; clean binary assignment; sponsor names classical ' +
-    'baseline (MIP/CP-SAT + heuristics) and defines success metric.\n' +
-    'BAD: constraint overhead dominates; equality constraints cause auxiliary variable explosion; ' +
-    'no baseline; no validation path.\n\n' +
-    'Portfolio optimisation:\n' +
-    'GOOD: discreteness essential (cardinality, turnover, lot sizes); constraints explicit; clear KPI; ' +
-    'backtesting plan exists.\n' +
-    'BAD: constraints subjective or constantly changing; baseline already strong with no improvement ' +
-    'target; no acceptance criteria.\n' +
-    'Benchmark start: 30-150 instruments with discrete constraints.\n\n' +
-    'Molecular simulation:\n' +
-    'GOOD: tight scientific scope; reduced Hamiltonian feasible; domain experts available; staged roadmap agreed.\n' +
-    'BAD: simulate everything with no reduction; no classical gold-standard comparison.\n\n' +
-    'ML / kernel methods:\n' +
-    'GOOD: narrow testable hybrid subroutine; measurable evaluation protocol; expectation is ' +
-    'benchmarking not production replacement.\n' +
-    'BAD: quantum replaces ML narrative; no baseline; no evaluation plan.\n\n' +
-    'Cryptography / PQC:\n' +
-    'Near-term value is PQC migration governance only. Flag and refuse conflation of quantum computing ' +
-    'with PQC -- these are completely separate problems.\n\n' +
-    'Sampling / Monte Carlo:\n' +
-    'GOOD: sampling is dominant cost; statistical success criteria exist.\n' +
-    'BAD: no verifiable metric; no classical baseline.\n\n' +
-    'QUBO FAILURE PATTERNS -- flag in dominant_blockers:\n' +
-    '- Heavy equality constraints or repeated exactly-one constraints at scale\n' +
-    '- Many if/then rules causing auxiliary variable explosion\n' +
-    '- Dense interaction graph post-encoding causing embedding blow-up\n' +
-    '- Penalty-dominated objectives: valid mathematically, meaningless commercially\n\n' +
-    'HARDWARE RULES (quantum_readiness_report only):\n' +
-    '- QUBO/Ising viable + new team: annealing/hybrid first (D-Wave Leap)\n' +
-    '- QAOA/variational: default simulator-only unless shallow circuit confirmed\n' +
-    '- New team or no baseline: simulator-only until baseline + reduction + validation exist\n' +
-    '- Key metric: two-qubit fidelity + effective depth survivability, NOT qubit count\n' +
-    '- Never recommend paid cloud hardware below suitability score 0.6\n\n' +
-    'SCORING:\n' +
-    '0.8+: all conditions met -- recommend structured pilot\n' +
-    '0.6-0.79: worth pilot with validation plan -- minimum for paid cloud run\n' +
-    '0.2-0.4: plausible class but baseline unclear -- needs reformulation\n' +
-    '0.05 or below: refuse immediately\n\n' +
+    '- "we already have a QUBO" but no penalty scaling explanation\n\n' +
+    'PROBLEM CLASS GUIDANCE:\n\n' +
+    'Portfolio optimisation: scientific_fit elevated when discreteness is essential (cardinality, turnover, lot sizes). ' +
+    'Benchmark start: 30–150 instruments. hardware_feasibility drops sharply above 300 assets. ' +
+    'advantage_claim_level is typically HYPOTHESISED for large-scale instances.\n\n' +
+    'Molecular simulation (VQE): scientific_fit 0.75–0.90 for reduced Hamiltonians. ' +
+    'hardware_feasibility 0.5–0.7 for small molecules on NISQ. ' +
+    'advantage_claim_level EXPERIMENTAL_SIGNAL for hydrogen chains, small molecules.\n\n' +
+    'Combinatorial optimisation: scientific_fit 0.4–0.7 depending on QUBO reducibility. ' +
+    'Flag in dominant_blockers: heavy equality constraints, auxiliary variable explosion, ' +
+    'penalty-dominated objectives.\n\n' +
+    'Cryptography/PQC: flag and refuse conflation of quantum computing with PQC migration. ' +
+    'These are completely separate problems.\n\n' +
+    'HARDWARE RULES:\n' +
+    '- QUBO/Ising viable: annealing/hybrid first (D-Wave Leap)\n' +
+    '- QAOA/variational: simulator-only unless shallow circuit confirmed\n' +
+    '- Never recommend paid cloud hardware with hardware_feasibility below 0.5\n\n' +
+    'ALWAYS populate baseline_question with:\n' +
+    '"What is your classical baseline today, and what metric must improve for this to matter?"\n\n' +
     'NEVER:\n' +
-    '- Use "quantum advantage" without problem-specific grounding\n' +
-    '- Promise quantum beats classical without a defined baseline\n' +
+    '- Assign PRODUCTION_VALIDATED without extraordinary published evidence\n' +
+    '- Assign COMMERCIALLY_RECOMMENDED_NOW to broad enterprise problem classes\n' +
     '- Recommend production hardware without prior simulator result\n' +
-    '- Present simulator wins as hardware wins\n' +
-    '- Recommend a pilot when motivation is investor optics';
+    '- Lower scientific_fit because commercial advantage is unproven';
 export function nowISO() {
     return new Date().toISOString();
 }

package/dist/constants.js.map

@@ -1 +1 @@
-{"version":3,"file":"constants.js","sourceRoot":"","sources":["../src/constants.ts"],"names":[],"mappings":"AAAA,MAAM,CAAC,MAAM,OAAO,GAAG,OAAO,CAAC;AAC/B,MAAM,CAAC,MAAM,qBAAqB,GAAG,EAAE,CAAC;AACxC,MAAM,CAAC,MAAM,eAAe,GAAG,KAAK,CAAC;AACrC,MAAM,CAAC,MAAM,eAAe,GAAG,CAAC,CAAC;AACjC,MAAM,CAAC,MAAM,YAAY,GAAG,yBAAyB,CAAC;AACtD,MAAM,CAAC,MAAM,eAAe,GAAG,gDAAgD,CAAC;AAChF,MAAM,CAAC,MAAM,sBAAsB,GAAG,gDAAgD,CAAC;AAEvF,MAAM,CAAC,MAAM,gBAAgB,GAC3B,6FAA6F;IAC7F,8GAA8G;IAC9G,qFAAqF;IACrF,mDAAmD;IACnD,6FAA6F;IAC7F,yCAAyC,CAAC;AAE5C,MAAM,CAAC,MAAM,aAAa,GACxB,+FAA+F;IAC/F,8FAA8F;IAC9F,kGAAkG;IAClG,wDAAwD;IACxD,2FAA2F;IAC3F,4DAA4D;IAC5D,wDAAwD;IACxD,oDAAoD;IACpD,+CAA+C;IAC/C,yDAAyD;IACzD,+DAA+D;IAC/D,gEAAgE;IAChE,iEAAiE;IACjE,8DAA8D;IAC9D,gDAAgD;IAChD,gCAAgC;IAChC,iEAAiE;IACjE,8CAA8C;IAC9C,+DAA+D;IAC/D,iEAAiE;IACjE,gEAAgE;IAChE,6FAA6F;IAC7F,4EAA4E;IAC5E,0BAA0B;IAC1B,+BAA+B;IAC/B,+FAA+F;IAC/F,kEAAkE;IAClE,+FAA+F;IAC/F,sCAAsC;IACtC,2BAA2B;IAC3B,oGAAoG;IACpG,4BAA4B;IAC5B,kGAAkG;IAClG,mCAAmC;IACnC,oEAAoE;IACpE,yBAAyB;IACzB,gHAAgH;IAChH,wFAAwF;IACxF,wBAAwB;IACxB,0FAA0F;IAC1F,4CAA4C;IAC5C,0EAA0E;IAC1E,uBAAuB;IACvB,oGAAoG;IACpG,yDAAyD;IACzD,2BAA2B;IAC3B,wEAAwE;IACxE,uDAAuD;IACvD,uDAAuD;IACvD,6EAA6E;IAC7E,6DAA6D;IAC7D,qEAAqE;IACrE,oFAAoF;IACpF,mDAAmD;IACnD,wEAAwE;IACxE,+EAA+E;IAC/E,2FAA2F;IAC3F,qFAAqF;IACrF,uEAAuE;IACvE,YAAY;IACZ,0DAA0D;IAC1D,4EAA4E;IAC5E,wEAAwE;IACxE,uCAAuC;IACvC,UAAU;IACV,gEAAgE;IAChE,gEAAgE;IAChE,kEAAkE;IAClE,6CAA6C;IAC7C,wDAAwD,CAAC;AAE3D,MAAM,UAAU,MAAM;IACpB,OAAO,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;AAClC,CAAC"}
+{"version":3,"file":"constants.js","sourceRoot":"","sources":["../src/constants.ts"],"names":[],"mappings":"AAAA,MAAM,CAAC,MAAM,OAAO,GAAG,QAAQ,CAAC;AAChC,MAAM,CAAC,MAAM,qBAAqB,GAAG,EAAE,CAAC;AACxC,MAAM,CAAC,MAAM,eAAe,GAAG,KAAK,CAAC;AACrC,MAAM,CAAC,MAAM,eAAe,GAAG,CAAC,CAAC;AACjC,MAAM,CAAC,MAAM,YAAY,GAAG,yBAAyB,CAAC;AACtD,MAAM,CAAC,MAAM,mBAAmB,GAAG,yBAAyB,CAAC;AAC7D,MAAM,CAAC,MAAM,eAAe,GAAG,gDAAgD,CAAC;AAChF,MAAM,CAAC,MAAM,sBAAsB,GAAG,gDAAgD,CAAC;AAEvF,MAAM,CAAC,MAAM,gBAAgB,GAC3B,6FAA6F;IAC7F,8GAA8G;IAC9G,qFAAqF;IACrF,mDAAmD;IACnD,6FAA6F;IAC7F,yCAAyC,CAAC;AAE5C,MAAM,CAAC,MAAM,aAAa,GACxB,2GAA2G;IAC3G,wGAAwG;IACxG,6HAA6H;IAE7H,yCAAyC;IAEzC,4DAA4D;IAE5D,+CAA+C;IAC/C,iDAAiD;IACjD,8GAA8G;IAC9G,0HAA0H;IAC1H,mCAAmC;IACnC,oFAAoF;IACpF,oEAAoE;IACpE,8EAA8E;IAC9E,wEAAwE;IAExE,wCAAwC;IACxC,mEAAmE;IACnE,8GAA8G;IAC9G,yDAAyD;IACzD,uEAAuE;IACvE,+EAA+E;IAC/E,6DAA6D;IAC7D,uDAAuD;IAEvD,uCAAuC;IACvC,mEAAmE;IACnE,8HAA8H;IAC9H,uFAAuF;IACvF,iFAAiF;IACjF,kEAAkE;IAClE,uEAAuE;IAEvE,wCAAwC;IACxC,sCAAsC;IACtC,oGAAoG;IACpG,2DAA2D;IAC3D,sDAAsD;IACtD,uEAAuE;IACvE,0DAA0D;IAC1D,yEAAyE;IAEzE,gIAAgI;IAEhI,uBAAuB;IACvB,0FAA0F;IAC1F,qFAAqF;IACrF,qEAAqE;IACrE,kHAAkH;IAClH,yEAAyE;IACzE,mGAAmG;IACnG,oGAAoG;IACpG,0GAA0G;IAC1G,mEAAmE;IACnE,6FAA6F;IAC7F,4GAA4G;IAE5G,sEAAsE;IACtE,2DAA2D;IAC3D,gGAAgG;IAChG,gGAAgG;IAChG,6FAA6F;IAC7F,wHAAwH;IAExH,kFAAkF;IAClF,mHAAmH;IACnH,0FAA0F;IAC1F,+FAA+F;IAC/F,mGAAmG;IACnG,gFAAgF;IAEhF,6CAA6C;IAC7C,8DAA8D;IAC9D,gDAAgD;IAChD,gCAAgC;IAChC,8CAA8C;IAC9C,+DAA+D;IAC/D,mEAAmE;IAEnE,6BAA6B;IAC7B,qHAAqH;IACrH,4FAA4F;IAC5F,gFAAgF;IAChF,iFAAiF;IACjF,4DAA4D;IAC5D,qFAAqF;IACrF,qFAAqF;IACrF,uFAAuF;IACvF,mCAAmC;IACnC,wFAAwF;IACxF,6CAA6C;IAE7C,mBAAmB;IACnB,6DAA6D;IAC7D,uEAAuE;IACvE,+EAA+E;IAE/E,2CAA2C;IAC3C,+FAA+F;IAE/F,UAAU;IACV,0EAA0E;IAC1E,6EAA6E;IAC7E,kEAAkE;IAClE,iEAAiE,CAAC;AAEpE,MAAM,UAAU,MAAM;IACpB,OAAO,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;AAClC,CAAC"}