agentic-qe 3.12.2 → 3.12.3

package/dist/cli/chunks/{chunk-VQU2JXTK.js → chunk-Y625LP6K.js}

@@ -1,5 +1,5 @@
-import{createRequire as __cr}from"module";const require=__cr(import.meta.url);if(process.argv.includes('--version')||process.argv.includes('-v')){console.log("3.12.2");process.exit(0)}
-import{a as T}from"./chunk-YY4KXDI6.js";import{c as S}from"./chunk-R5HC6JVJ.js";import{b as C,d as P}from"./chunk-CKFZOPZN.js";import{a as b,c as $}from"./chunk-X47GHZY3.js";import{a as w,d as M}from"./chunk-4KEOBDEI.js";$();M();import{randomUUID as A}from"crypto";var x={topology:"hierarchical",maxConcurrentSkills:10,maxConcurrentModels:3,timeout:3e5,continueOnFailure:!0,retry:{maxRetries:2,retryDelayMs:1e3}},R=class{config;learner;skillValidator=null;workers=new Map;taskQueue=[];isRunning=!1;constructor(e,t){this.config={...x,...e},this.learner=t}setSkillValidator(e){this.skillValidator=e}async validateSkillsParallel(e,t,i){let s=Date.now(),a=i?.trustTier??3,n=i?.validationLevel??"eval",c=this.determineTopology(e.length,t.length),r=this.createValidationTasks(e,t,a,n),o=await this.executeTasksParallel(r,c),l=new Map;for(let u of o){let p=l.get(u.skill)||[];p.push(u),l.set(u.skill,p)}return await this.recordOutcomesToLearner(o,a,n),l}async validateSkillCrossModel(e,t,i){return(await this.validateSkillsParallel([e],t,i)).get(e)||[]}getSummary(e){let t=[],i=new Map,s=new Map;for(let[c,r]of e){i.set(c,r);for(let o of r){t.push(o);let l=s.get(o.model)||[];l.push(o),s.set(o.model,l)}}let a=t.filter(c=>c.errors.length===0&&c.evalPassRate>=.9).length,n=t.reduce((c,r)=>c+r.durationMs,0);return{totalSkills:i.size,totalModels:s.size,successCount:a,failureCount:t.length-a,overallPassRate:t.length>0?a/t.length:0,totalDurationMs:n,avgDurationMs:t.length>0?n/t.length:0,topology:this.config.topology,bySkill:i,byModel:s,results:t}}determineTopology(e,t){if(this.config.topology){let i=e*t;return i>20&&this.config.topology==="mesh"&&console.warn(`[SwarmSkillValidator] Large workload (${i} tasks) may be more efficient with hierarchical topology`),this.config.topology}return e>5||e>t*2?"hierarchical":t>5&&e<=3?"mesh":"hierarchical"}createValidationTasks(e,t,i,s){let a=[];for(let n of e)for(let c of t)a.push({id:`${n}-${c}-${Date.now()}-${A().slice(0,8)}`,skill:n,model:c,trustTier:i,validationLevel:s,timeout:this.config.timeout,retryCount:0});return a}async executeTasksParallel(e,t){this.isRunning=!0,this.taskQueue=[...e];let i=[],s=t==="hierarchical"?this.config.maxConcurrentSkills:Math.min(this.config.maxConcurrentSkills*this.config.maxConcurrentModels,e.length),a=[];for(let n=0;n<Math.min(s,e.length);n++){let c=`worker-${n}`;this.workers.set(c,{id:c,task:null,status:"idle"}),a.push(this.runWorker(c,i))}return await Promise.all(a),this.isRunning=!1,this.workers.clear(),i}async runWorker(e,t){let i=this.workers.get(e);if(i){for(;this.taskQueue.length>0&&this.isRunning;){let s=this.taskQueue.shift();if(!s)break;i.task=s,i.status="running",i.startTime=Date.now();try{let a=await this.executeValidationTask(s);t.push(a),i.result=a,i.status="completed"}catch(a){let n=this.createErrorResult(s,a);this.config.retry&&s.retryCount<this.config.retry.maxRetries&&this.config.continueOnFailure?(s.retryCount++,this.taskQueue.push(s),await this.delay(this.config.retry.retryDelayMs)):(t.push(n),i.result=n,i.status="failed")}}i.task=null,i.status="idle"}}async executeValidationTask(e){let t=Date.now();if(!this.skillValidator)return this.createSimulatedResult(e,t);let i=new Promise((n,c)=>{setTimeout(()=>c(new Error(`Validation timeout after ${e.timeout}ms`)),e.timeout)}),s=await Promise.race([this.skillValidator(e.skill,e.model,{trustTier:e.trustTier,validationLevel:e.validationLevel,timeout:e.timeout}),i]),a=Date.now()-t;return this.outcomeToResult(s,e,a)}outcomeToResult(e,t,i){let s=e.testCaseResults.filter(c=>c.passed).length,a=e.testCaseResults.length,n=a>0?s/a:e.score;return{skill:t.skill,model:t.model,schemaValid:e.validationLevel!=="schema"||e.passed,validatorPassed:e.validationLevel!=="validator"||e.passed,evalPassRate:n,durationMs:i,errors:e.passed?[]:["Validation failed"],trustTier:t.trustTier,validationLevel:t.validationLevel,retryCount:t.retryCount,timestamp:new Date}}createSimulatedResult(e,t){let i=Date.now()-t+w()*100,s=.85+w()*.15;return{skill:e.skill,model:e.model,schemaValid:!0,validatorPassed:s>.8,evalPassRate:s,durationMs:i,errors:s>.8?[]:["Simulated validation failure"],trustTier:e.trustTier,validationLevel:e.validationLevel,retryCount:e.retryCount,timestamp:new Date}}createErrorResult(e,t){let i=b(t);return{skill:e.skill,model:e.model,schemaValid:!1,validatorPassed:!1,evalPassRate:0,durationMs:0,errors:[i],trustTier:e.trustTier,validationLevel:e.validationLevel,retryCount:e.retryCount,timestamp:new Date}}async recordOutcomesToLearner(e,t,i){for(let s of e){let a={skillName:s.skill,trustTier:s.trustTier||t,validationLevel:s.validationLevel||i,model:s.model,passed:s.errors.length===0&&s.evalPassRate>=.9,score:s.evalPassRate,testCaseResults:this.createTestCaseResults(s),timestamp:s.timestamp,runId:`swarm-${Date.now()}`,metadata:{duration:s.durationMs,retryCount:s.retryCount}};await this.learner.recordValidationOutcome(a)}}createTestCaseResults(e){return[{testId:`${e.skill}-${e.model}-aggregate`,passed:e.errors.length===0,expectedPatterns:["valid-output"],actualPatterns:e.errors.length===0?["valid-output"]:[],reasoningQuality:e.evalPassRate,executionTimeMs:e.durationMs,category:"swarm-validation",priority:"high",error:e.errors.length>0?e.errors.join("; "):void 0}]}delay(e){return new Promise(t=>setTimeout(t,e))}getConfig(){return{...this.config}}isValidationRunning(){return this.isRunning}getWorkerStatus(){return Array.from(this.workers.values())}cancel(){this.isRunning=!1,this.taskQueue=[]}};function D(y,e){return new R(y,e)}var F=["security-testing","accessibility-testing","api-testing","performance-testing","visual-regression-testing","mutation-testing","contract-testing","chaos-testing","compliance-testing","penetration-testing"],I=["claude-sonnet","claude-haiku","claude-opus"];P();import{readFileSync as E,writeFileSync as L,existsSync as N}from"fs";var O={varianceThreshold:.04,regressionThreshold:.1,minSamples:3,autoUpdateManifest:!1},k=class{constructor(e,t,i={}){this.learner=e;this.manifestPath=t;this.config={...O,...i}}learner;manifestPath;config;async aggregateResults(e){let t=`agg-${Date.now()}`,i=new Date,s=this.buildSkillResultsMap(e),a=this.calculateSummary(s,e),n=await this.detectCrossModelAnomalies(e),c=await this.detectRegressions(e,this.config.regressionThreshold),r=this.generateRecommendations(s,n,c),o={timestamp:i,runId:t,summary:a,skillResults:s,crossModelAnalysis:n,regressions:c,recommendations:r,metadata:{version:"1.0.0",environment:e[0]?.metadata?.environment,generatedBy:"ValidationResultAggregator",inputs:{runIds:e.map(l=>l.runId),models:[...new Set(e.map(l=>l.model))]}}};return this.config.autoUpdateManifest&&await this.updateManifest(o),o}buildSkillResultsMap(e){let t=new Map;for(let i of e)for(let s of i.outcomes){let a=s.skillName;t.has(a)||t.set(a,{skill:a,trustTier:s.trustTier,passRateByModel:new Map,avgPassRate:0,schemaValid:!0,validatorPassed:!0,evalPassed:!0,issues:[],executionTimeMs:0,testCount:0,passedTests:0,failedTests:0});let n=t.get(a),c=s.testCaseResults.filter(r=>r.passed).length/(s.testCaseResults.length||1);n.passRateByModel.set(i.model,c),n.testCount+=s.testCaseResults.length,n.passedTests+=s.testCaseResults.filter(r=>r.passed).length,n.failedTests+=s.testCaseResults.filter(r=>!r.passed).length,n.executionTimeMs+=s.metadata?.duration||0,s.validationLevel==="schema"&&!s.passed&&(n.schemaValid=!1),s.validationLevel==="validator"&&!s.passed&&(n.validatorPassed=!1),s.validationLevel==="eval"&&!s.passed&&(n.evalPassed=!1);for(let r of s.testCaseResults)!r.passed&&r.error&&n.issues.push({skill:a,model:i.model,severity:r.priority==="critical"?"critical":r.priority==="high"?"high":r.priority==="medium"?"medium":"low",type:this.categorizeIssueType(s.validationLevel,r.error),message:r.error,testId:r.testId})}for(let i of t.values()){let s=Array.from(i.passRateByModel.values());i.avgPassRate=s.length>0?s.reduce((a,n)=>a+n,0)/s.length:0}return t}categorizeIssueType(e,t){return t.toLowerCase().includes("timeout")?"timeout":e==="schema"?"schema_failure":e==="validator"?"validator_failure":e==="eval"?"eval_failure":"error"}calculateSummary(e,t){let i=Array.from(e.values()),s=i.filter(n=>n.avgPassRate>=.9).length,a=i.length>0?i.reduce((n,c)=>n+c.avgPassRate,0)/i.length:0;return{totalSkills:i.length,passedSkills:s,failedSkills:i.length-s,avgPassRate:a,totalDurationMs:t.reduce((n,c)=>n+c.durationMs,0),totalTests:i.reduce((n,c)=>n+c.testCount,0),passedTests:i.reduce((n,c)=>n+c.passedTests,0),failedTests:i.reduce((n,c)=>n+c.failedTests,0),modelsUsed:[...new Set(t.map(n=>n.model))]}}async detectCrossModelAnomalies(e){let t=new Map,i=new Map,s=new Map;for(let d of e){t.has(d.model)||(t.set(d.model,[]),i.set(d.model,{totalTests:0,passedTests:0,skillCount:0}));for(let g of d.outcomes){let h=g.testCaseResults.filter(m=>m.passed).length/(g.testCaseResults.length||1);t.get(d.model).push(h);let f=i.get(d.model);f.totalTests+=g.testCaseResults.length,f.passedTests+=g.testCaseResults.filter(m=>m.passed).length,f.skillCount++,s.has(g.skillName)||s.set(g.skillName,{passRates:[],models:[]}),s.get(g.skillName).passRates.push(h),s.get(g.skillName).models.push(d.model)}}let a=[];for(let d of t.values())a.push(...d);let n=a.length>0?a.reduce((d,g)=>d+g,0)/a.length:0,c=a.length>1?a.reduce((d,g)=>d+Math.pow(g-n,2),0)/a.length:0,r=Math.sqrt(c),o=[],l=new Map;for(let[d,g]of t.entries()){let h=g.reduce((v,V)=>v+V,0)/g.length,f=h-n,m=i.get(d);l.set(d,{avgPassRate:h,skillCount:m.skillCount,totalTests:m.totalTests}),Math.abs(f)>this.config.varianceThreshold*2&&o.push({model:d,type:f<0?"low_performance":"high_variance",description:`Model ${d} has ${(f*100).toFixed(1)}% difference from average`,passRate:h,avgPassRate:n,deviation:Math.abs(f)})}let u=[],p=[];for(let[d,g]of s.entries()){if(g.passRates.length<2)continue;let h=g.passRates.reduce((m,v)=>m+v,0)/g.passRates.length;g.passRates.reduce((m,v)=>m+Math.pow(v-h,2),0)/g.passRates.length<this.config.varianceThreshold?u.push(d):p.push(d)}return{variance:c,stdDeviation:r,anomalies:o,consistentSkills:u,inconsistentSkills:p,modelPerformance:l}}async detectRegressions(e,t){let i=[],s=new Set;for(let a of e)for(let n of a.outcomes){let c=`${n.skillName}-${a.model}`;if(s.has(c))continue;s.add(c);let r=await this.learner.getValidationTrends(n.skillName);if(!r)continue;let o=n.testCaseResults.filter(p=>p.passed).length/(n.testCaseResults.length||1),l=r.recentPassRate,u=l-o;u>=t&&i.push({skill:n.skillName,model:a.model,previousPassRate:l,currentPassRate:o,regressionAmount:u,trend:r.overall,possibleCauses:this.analyzePossibleCauses(n,u),severity:this.categorizeRegressionSeverity(u)})}return i.sort((a,n)=>{let c={critical:0,high:1,medium:2,low:3},r=c[a.severity]-c[n.severity];return r!==0?r:n.regressionAmount-a.regressionAmount})}analyzePossibleCauses(e,t){let i=[],s=e.testCaseResults.filter(c=>!c.passed);if(s.length>0){let c=[...new Set(s.map(r=>r.category).filter(Boolean))];c.length>0&&i.push(`Failures in categories: ${c.join(", ")}`)}let a=s.filter(c=>c.priority==="critical");a.length>0&&i.push(`${a.length} critical test(s) failing`);let n=e.testCaseResults.reduce((c,r)=>c+r.reasoningQuality,0)/(e.testCaseResults.length||1);return n<.7&&i.push(`Low reasoning quality score: ${(n*100).toFixed(1)}%`),t>.3&&i.push("Possible model behavior change or prompt drift"),i.length===0&&i.push("No specific cause identified - review test details"),i}categorizeRegressionSeverity(e){return e>=.5?"critical":e>=.3?"high":e>=.15?"medium":"low"}generateRecommendations(e,t,i){let s=[];if(t.inconsistentSkills.length>0&&s.push(`Review ${t.inconsistentSkills.length} skills with inconsistent cross-model behavior: `+t.inconsistentSkills.slice(0,3).join(", ")+(t.inconsistentSkills.length>3?` and ${t.inconsistentSkills.length-3} more`:"")),t.anomalies.some(r=>r.type==="low_performance")){let r=t.anomalies.filter(o=>o.type==="low_performance").map(o=>o.model);s.push(`Investigate low performance on model(s): ${r.join(", ")}`)}let a=i.filter(r=>r.severity==="critical");a.length>0&&s.push(`URGENT: ${a.length} critical regression(s) detected - immediate review required`);let n=Array.from(e.values()).filter(r=>r.avgPassRate<.9);if(n.length>0){let r=n.filter(l=>!l.schemaValid),o=n.filter(l=>!l.validatorPassed);r.length>0&&s.push(`Fix schema validation for: ${r.map(l=>l.skill).slice(0,3).join(", ")}`),o.length>0&&s.push(`Review validator scripts for: ${o.map(l=>l.skill).slice(0,3).join(", ")}`)}return Array.from(e.values()).reduce((r,o)=>r+o.avgPassRate,0)/e.size<.8&&s.push("Overall pass rate below 80% - consider reviewing skill definitions and test expectations"),s.length===0&&s.push("All validations passing - no immediate action required"),s}async updateManifest(e){if(!N(this.manifestPath))throw new Error(`Manifest file not found: ${this.manifestPath}`);let t=C(E(this.manifestPath,"utf-8"));for(let[s,a]of e.skillResults.entries())if(t.skills&&t.skills[s]){let n=t.skills[s];n.validation||(n.validation={}),n.validation.passRate=a.avgPassRate,n.validation.lastValidated=e.timestamp.toISOString(),n.validation.status=a.avgPassRate>=.9?"passing":"failing",n.validation.passRateByModel=Object.fromEntries(a.passRateByModel)}let i=Array.from(e.skillResults.values());t.validationStatus={passing:i.filter(s=>s.avgPassRate>=.9).length,failing:i.filter(s=>s.avgPassRate<.9).length,unknown:0,skipped:t.summary?.tier0||0},t.generatedAt=new Date().toISOString(),t.lastValidationRun={runId:e.runId,timestamp:e.timestamp.toISOString(),avgPassRate:e.summary.avgPassRate,modelsUsed:e.summary.modelsUsed},L(this.manifestPath,JSON.stringify(t,null,2))}generateMarkdownReport(e){let{summary:t,crossModelAnalysis:i,regressions:s,recommendations:a,skillResults:n}=e,c=(t.avgPassRate>=.9||t.avgPassRate>=.7,"`"),r=[];if(r.push(`# Validation Report
+import{createRequire as __cr}from"module";const require=__cr(import.meta.url);if(process.argv.includes('--version')||process.argv.includes('-v')){console.log("3.12.3");process.exit(0)}
+import{a as T}from"./chunk-G5OGKVLO.js";import{c as S}from"./chunk-4EHW4TNA.js";import{b as C,d as P}from"./chunk-OBL6W4Z5.js";import{a as b,c as $}from"./chunk-2LNDU7AW.js";import{a as w,d as M}from"./chunk-YUDGQZJN.js";$();M();import{randomUUID as A}from"crypto";var x={topology:"hierarchical",maxConcurrentSkills:10,maxConcurrentModels:3,timeout:3e5,continueOnFailure:!0,retry:{maxRetries:2,retryDelayMs:1e3}},R=class{config;learner;skillValidator=null;workers=new Map;taskQueue=[];isRunning=!1;constructor(e,t){this.config={...x,...e},this.learner=t}setSkillValidator(e){this.skillValidator=e}async validateSkillsParallel(e,t,i){let s=Date.now(),a=i?.trustTier??3,n=i?.validationLevel??"eval",c=this.determineTopology(e.length,t.length),r=this.createValidationTasks(e,t,a,n),o=await this.executeTasksParallel(r,c),l=new Map;for(let u of o){let p=l.get(u.skill)||[];p.push(u),l.set(u.skill,p)}return await this.recordOutcomesToLearner(o,a,n),l}async validateSkillCrossModel(e,t,i){return(await this.validateSkillsParallel([e],t,i)).get(e)||[]}getSummary(e){let t=[],i=new Map,s=new Map;for(let[c,r]of e){i.set(c,r);for(let o of r){t.push(o);let l=s.get(o.model)||[];l.push(o),s.set(o.model,l)}}let a=t.filter(c=>c.errors.length===0&&c.evalPassRate>=.9).length,n=t.reduce((c,r)=>c+r.durationMs,0);return{totalSkills:i.size,totalModels:s.size,successCount:a,failureCount:t.length-a,overallPassRate:t.length>0?a/t.length:0,totalDurationMs:n,avgDurationMs:t.length>0?n/t.length:0,topology:this.config.topology,bySkill:i,byModel:s,results:t}}determineTopology(e,t){if(this.config.topology){let i=e*t;return i>20&&this.config.topology==="mesh"&&console.warn(`[SwarmSkillValidator] Large workload (${i} tasks) may be more efficient with hierarchical topology`),this.config.topology}return e>5||e>t*2?"hierarchical":t>5&&e<=3?"mesh":"hierarchical"}createValidationTasks(e,t,i,s){let a=[];for(let n of e)for(let c of t)a.push({id:`${n}-${c}-${Date.now()}-${A().slice(0,8)}`,skill:n,model:c,trustTier:i,validationLevel:s,timeout:this.config.timeout,retryCount:0});return a}async executeTasksParallel(e,t){this.isRunning=!0,this.taskQueue=[...e];let i=[],s=t==="hierarchical"?this.config.maxConcurrentSkills:Math.min(this.config.maxConcurrentSkills*this.config.maxConcurrentModels,e.length),a=[];for(let n=0;n<Math.min(s,e.length);n++){let c=`worker-${n}`;this.workers.set(c,{id:c,task:null,status:"idle"}),a.push(this.runWorker(c,i))}return await Promise.all(a),this.isRunning=!1,this.workers.clear(),i}async runWorker(e,t){let i=this.workers.get(e);if(i){for(;this.taskQueue.length>0&&this.isRunning;){let s=this.taskQueue.shift();if(!s)break;i.task=s,i.status="running",i.startTime=Date.now();try{let a=await this.executeValidationTask(s);t.push(a),i.result=a,i.status="completed"}catch(a){let n=this.createErrorResult(s,a);this.config.retry&&s.retryCount<this.config.retry.maxRetries&&this.config.continueOnFailure?(s.retryCount++,this.taskQueue.push(s),await this.delay(this.config.retry.retryDelayMs)):(t.push(n),i.result=n,i.status="failed")}}i.task=null,i.status="idle"}}async executeValidationTask(e){let t=Date.now();if(!this.skillValidator)return this.createSimulatedResult(e,t);let i=new Promise((n,c)=>{setTimeout(()=>c(new Error(`Validation timeout after ${e.timeout}ms`)),e.timeout)}),s=await Promise.race([this.skillValidator(e.skill,e.model,{trustTier:e.trustTier,validationLevel:e.validationLevel,timeout:e.timeout}),i]),a=Date.now()-t;return this.outcomeToResult(s,e,a)}outcomeToResult(e,t,i){let s=e.testCaseResults.filter(c=>c.passed).length,a=e.testCaseResults.length,n=a>0?s/a:e.score;return{skill:t.skill,model:t.model,schemaValid:e.validationLevel!=="schema"||e.passed,validatorPassed:e.validationLevel!=="validator"||e.passed,evalPassRate:n,durationMs:i,errors:e.passed?[]:["Validation failed"],trustTier:t.trustTier,validationLevel:t.validationLevel,retryCount:t.retryCount,timestamp:new Date}}createSimulatedResult(e,t){let i=Date.now()-t+w()*100,s=.85+w()*.15;return{skill:e.skill,model:e.model,schemaValid:!0,validatorPassed:s>.8,evalPassRate:s,durationMs:i,errors:s>.8?[]:["Simulated validation failure"],trustTier:e.trustTier,validationLevel:e.validationLevel,retryCount:e.retryCount,timestamp:new Date}}createErrorResult(e,t){let i=b(t);return{skill:e.skill,model:e.model,schemaValid:!1,validatorPassed:!1,evalPassRate:0,durationMs:0,errors:[i],trustTier:e.trustTier,validationLevel:e.validationLevel,retryCount:e.retryCount,timestamp:new Date}}async recordOutcomesToLearner(e,t,i){for(let s of e){let a={skillName:s.skill,trustTier:s.trustTier||t,validationLevel:s.validationLevel||i,model:s.model,passed:s.errors.length===0&&s.evalPassRate>=.9,score:s.evalPassRate,testCaseResults:this.createTestCaseResults(s),timestamp:s.timestamp,runId:`swarm-${Date.now()}`,metadata:{duration:s.durationMs,retryCount:s.retryCount}};await this.learner.recordValidationOutcome(a)}}createTestCaseResults(e){return[{testId:`${e.skill}-${e.model}-aggregate`,passed:e.errors.length===0,expectedPatterns:["valid-output"],actualPatterns:e.errors.length===0?["valid-output"]:[],reasoningQuality:e.evalPassRate,executionTimeMs:e.durationMs,category:"swarm-validation",priority:"high",error:e.errors.length>0?e.errors.join("; "):void 0}]}delay(e){return new Promise(t=>setTimeout(t,e))}getConfig(){return{...this.config}}isValidationRunning(){return this.isRunning}getWorkerStatus(){return Array.from(this.workers.values())}cancel(){this.isRunning=!1,this.taskQueue=[]}};function D(y,e){return new R(y,e)}var F=["security-testing","accessibility-testing","api-testing","performance-testing","visual-regression-testing","mutation-testing","contract-testing","chaos-testing","compliance-testing","penetration-testing"],I=["claude-sonnet","claude-haiku","claude-opus"];P();import{readFileSync as E,writeFileSync as L,existsSync as N}from"fs";var O={varianceThreshold:.04,regressionThreshold:.1,minSamples:3,autoUpdateManifest:!1},k=class{constructor(e,t,i={}){this.learner=e;this.manifestPath=t;this.config={...O,...i}}learner;manifestPath;config;async aggregateResults(e){let t=`agg-${Date.now()}`,i=new Date,s=this.buildSkillResultsMap(e),a=this.calculateSummary(s,e),n=await this.detectCrossModelAnomalies(e),c=await this.detectRegressions(e,this.config.regressionThreshold),r=this.generateRecommendations(s,n,c),o={timestamp:i,runId:t,summary:a,skillResults:s,crossModelAnalysis:n,regressions:c,recommendations:r,metadata:{version:"1.0.0",environment:e[0]?.metadata?.environment,generatedBy:"ValidationResultAggregator",inputs:{runIds:e.map(l=>l.runId),models:[...new Set(e.map(l=>l.model))]}}};return this.config.autoUpdateManifest&&await this.updateManifest(o),o}buildSkillResultsMap(e){let t=new Map;for(let i of e)for(let s of i.outcomes){let a=s.skillName;t.has(a)||t.set(a,{skill:a,trustTier:s.trustTier,passRateByModel:new Map,avgPassRate:0,schemaValid:!0,validatorPassed:!0,evalPassed:!0,issues:[],executionTimeMs:0,testCount:0,passedTests:0,failedTests:0});let n=t.get(a),c=s.testCaseResults.filter(r=>r.passed).length/(s.testCaseResults.length||1);n.passRateByModel.set(i.model,c),n.testCount+=s.testCaseResults.length,n.passedTests+=s.testCaseResults.filter(r=>r.passed).length,n.failedTests+=s.testCaseResults.filter(r=>!r.passed).length,n.executionTimeMs+=s.metadata?.duration||0,s.validationLevel==="schema"&&!s.passed&&(n.schemaValid=!1),s.validationLevel==="validator"&&!s.passed&&(n.validatorPassed=!1),s.validationLevel==="eval"&&!s.passed&&(n.evalPassed=!1);for(let r of s.testCaseResults)!r.passed&&r.error&&n.issues.push({skill:a,model:i.model,severity:r.priority==="critical"?"critical":r.priority==="high"?"high":r.priority==="medium"?"medium":"low",type:this.categorizeIssueType(s.validationLevel,r.error),message:r.error,testId:r.testId})}for(let i of t.values()){let s=Array.from(i.passRateByModel.values());i.avgPassRate=s.length>0?s.reduce((a,n)=>a+n,0)/s.length:0}return t}categorizeIssueType(e,t){return t.toLowerCase().includes("timeout")?"timeout":e==="schema"?"schema_failure":e==="validator"?"validator_failure":e==="eval"?"eval_failure":"error"}calculateSummary(e,t){let i=Array.from(e.values()),s=i.filter(n=>n.avgPassRate>=.9).length,a=i.length>0?i.reduce((n,c)=>n+c.avgPassRate,0)/i.length:0;return{totalSkills:i.length,passedSkills:s,failedSkills:i.length-s,avgPassRate:a,totalDurationMs:t.reduce((n,c)=>n+c.durationMs,0),totalTests:i.reduce((n,c)=>n+c.testCount,0),passedTests:i.reduce((n,c)=>n+c.passedTests,0),failedTests:i.reduce((n,c)=>n+c.failedTests,0),modelsUsed:[...new Set(t.map(n=>n.model))]}}async detectCrossModelAnomalies(e){let t=new Map,i=new Map,s=new Map;for(let d of e){t.has(d.model)||(t.set(d.model,[]),i.set(d.model,{totalTests:0,passedTests:0,skillCount:0}));for(let g of d.outcomes){let h=g.testCaseResults.filter(m=>m.passed).length/(g.testCaseResults.length||1);t.get(d.model).push(h);let f=i.get(d.model);f.totalTests+=g.testCaseResults.length,f.passedTests+=g.testCaseResults.filter(m=>m.passed).length,f.skillCount++,s.has(g.skillName)||s.set(g.skillName,{passRates:[],models:[]}),s.get(g.skillName).passRates.push(h),s.get(g.skillName).models.push(d.model)}}let a=[];for(let d of t.values())a.push(...d);let n=a.length>0?a.reduce((d,g)=>d+g,0)/a.length:0,c=a.length>1?a.reduce((d,g)=>d+Math.pow(g-n,2),0)/a.length:0,r=Math.sqrt(c),o=[],l=new Map;for(let[d,g]of t.entries()){let h=g.reduce((v,V)=>v+V,0)/g.length,f=h-n,m=i.get(d);l.set(d,{avgPassRate:h,skillCount:m.skillCount,totalTests:m.totalTests}),Math.abs(f)>this.config.varianceThreshold*2&&o.push({model:d,type:f<0?"low_performance":"high_variance",description:`Model ${d} has ${(f*100).toFixed(1)}% difference from average`,passRate:h,avgPassRate:n,deviation:Math.abs(f)})}let u=[],p=[];for(let[d,g]of s.entries()){if(g.passRates.length<2)continue;let h=g.passRates.reduce((m,v)=>m+v,0)/g.passRates.length;g.passRates.reduce((m,v)=>m+Math.pow(v-h,2),0)/g.passRates.length<this.config.varianceThreshold?u.push(d):p.push(d)}return{variance:c,stdDeviation:r,anomalies:o,consistentSkills:u,inconsistentSkills:p,modelPerformance:l}}async detectRegressions(e,t){let i=[],s=new Set;for(let a of e)for(let n of a.outcomes){let c=`${n.skillName}-${a.model}`;if(s.has(c))continue;s.add(c);let r=await this.learner.getValidationTrends(n.skillName);if(!r)continue;let o=n.testCaseResults.filter(p=>p.passed).length/(n.testCaseResults.length||1),l=r.recentPassRate,u=l-o;u>=t&&i.push({skill:n.skillName,model:a.model,previousPassRate:l,currentPassRate:o,regressionAmount:u,trend:r.overall,possibleCauses:this.analyzePossibleCauses(n,u),severity:this.categorizeRegressionSeverity(u)})}return i.sort((a,n)=>{let c={critical:0,high:1,medium:2,low:3},r=c[a.severity]-c[n.severity];return r!==0?r:n.regressionAmount-a.regressionAmount})}analyzePossibleCauses(e,t){let i=[],s=e.testCaseResults.filter(c=>!c.passed);if(s.length>0){let c=[...new Set(s.map(r=>r.category).filter(Boolean))];c.length>0&&i.push(`Failures in categories: ${c.join(", ")}`)}let a=s.filter(c=>c.priority==="critical");a.length>0&&i.push(`${a.length} critical test(s) failing`);let n=e.testCaseResults.reduce((c,r)=>c+r.reasoningQuality,0)/(e.testCaseResults.length||1);return n<.7&&i.push(`Low reasoning quality score: ${(n*100).toFixed(1)}%`),t>.3&&i.push("Possible model behavior change or prompt drift"),i.length===0&&i.push("No specific cause identified - review test details"),i}categorizeRegressionSeverity(e){return e>=.5?"critical":e>=.3?"high":e>=.15?"medium":"low"}generateRecommendations(e,t,i){let s=[];if(t.inconsistentSkills.length>0&&s.push(`Review ${t.inconsistentSkills.length} skills with inconsistent cross-model behavior: `+t.inconsistentSkills.slice(0,3).join(", ")+(t.inconsistentSkills.length>3?` and ${t.inconsistentSkills.length-3} more`:"")),t.anomalies.some(r=>r.type==="low_performance")){let r=t.anomalies.filter(o=>o.type==="low_performance").map(o=>o.model);s.push(`Investigate low performance on model(s): ${r.join(", ")}`)}let a=i.filter(r=>r.severity==="critical");a.length>0&&s.push(`URGENT: ${a.length} critical regression(s) detected - immediate review required`);let n=Array.from(e.values()).filter(r=>r.avgPassRate<.9);if(n.length>0){let r=n.filter(l=>!l.schemaValid),o=n.filter(l=>!l.validatorPassed);r.length>0&&s.push(`Fix schema validation for: ${r.map(l=>l.skill).slice(0,3).join(", ")}`),o.length>0&&s.push(`Review validator scripts for: ${o.map(l=>l.skill).slice(0,3).join(", ")}`)}return Array.from(e.values()).reduce((r,o)=>r+o.avgPassRate,0)/e.size<.8&&s.push("Overall pass rate below 80% - consider reviewing skill definitions and test expectations"),s.length===0&&s.push("All validations passing - no immediate action required"),s}async updateManifest(e){if(!N(this.manifestPath))throw new Error(`Manifest file not found: ${this.manifestPath}`);let t=C(E(this.manifestPath,"utf-8"));for(let[s,a]of e.skillResults.entries())if(t.skills&&t.skills[s]){let n=t.skills[s];n.validation||(n.validation={}),n.validation.passRate=a.avgPassRate,n.validation.lastValidated=e.timestamp.toISOString(),n.validation.status=a.avgPassRate>=.9?"passing":"failing",n.validation.passRateByModel=Object.fromEntries(a.passRateByModel)}let i=Array.from(e.skillResults.values());t.validationStatus={passing:i.filter(s=>s.avgPassRate>=.9).length,failing:i.filter(s=>s.avgPassRate<.9).length,unknown:0,skipped:t.summary?.tier0||0},t.generatedAt=new Date().toISOString(),t.lastValidationRun={runId:e.runId,timestamp:e.timestamp.toISOString(),avgPassRate:e.summary.avgPassRate,modelsUsed:e.summary.modelsUsed},L(this.manifestPath,JSON.stringify(t,null,2))}generateMarkdownReport(e){let{summary:t,crossModelAnalysis:i,regressions:s,recommendations:a,skillResults:n}=e,c=(t.avgPassRate>=.9||t.avgPassRate>=.7,"`"),r=[];if(r.push(`# Validation Report
 
 > ${c} Generated: ${e.timestamp.toISOString()}
 > Run ID: ${e.runId}

package/dist/cli/chunks/{chunk-56MXYSYY.js → chunk-Y6XNISHP.js}

@@ -1,2 +1,2 @@
-import{createRequire as __cr}from"module";const require=__cr(import.meta.url);if(process.argv.includes('--version')||process.argv.includes('-v')){console.log("3.12.2");process.exit(0)}
-import{i as e}from"./chunk-ZBYMF2QJ.js";function g(t,n,r,a){return{id:e(),type:t,timestamp:new Date,source:n,correlationId:a,payload:r}}var i={TestGenerated:"test-generation.TestGenerated",TestSuiteCreated:"test-generation.TestSuiteCreated",PatternLearned:"test-generation.PatternLearned",GenerationFailed:"test-generation.GenerationFailed"},s={TestRunStarted:"test-execution.TestRunStarted",TestRunCompleted:"test-execution.TestRunCompleted",FlakyTestDetected:"test-execution.FlakyTestDetected",RetryTriggered:"test-execution.RetryTriggered"},o={CoverageReportCreated:"coverage-analysis.CoverageReportCreated",CoverageGapDetected:"coverage-analysis.CoverageGapDetected",RiskZoneIdentified:"coverage-analysis.RiskZoneIdentified"},d={QualityGateEvaluated:"quality-assessment.QualityGateEvaluated",DeploymentApproved:"quality-assessment.DeploymentApproved",DeploymentBlocked:"quality-assessment.DeploymentBlocked"},l={DefectPredicted:"defect-intelligence.DefectPredicted",RootCauseIdentified:"defect-intelligence.RootCauseIdentified",RegressionRiskAnalyzed:"defect-intelligence.RegressionRiskAnalyzed"},c={KnowledgeGraphUpdated:"code-intelligence.KnowledgeGraphUpdated",ImpactAnalysisCompleted:"code-intelligence.ImpactAnalysisCompleted",SemanticSearchCompleted:"code-intelligence.SemanticSearchCompleted",C4DiagramsGenerated:"code-intelligence.C4DiagramsGenerated"},m={VulnerabilityDetected:"security-compliance.VulnerabilityDetected",ComplianceValidated:"security-compliance.ComplianceValidated",SecurityAuditCompleted:"security-compliance.SecurityAuditCompleted"},p={PatternConsolidated:"learning-optimization.PatternConsolidated",TransferCompleted:"learning-optimization.TransferCompleted",OptimizationApplied:"learning-optimization.OptimizationApplied",DreamCycleCompleted:"learning-optimization.dream.completed"},y={...i,...s,...o,...d,...l,...c,...m,...p};export{g as a,i as b,s as c,o as d,d as e,l as f,c as g,m as h,p as i};
+import{createRequire as __cr}from"module";const require=__cr(import.meta.url);if(process.argv.includes('--version')||process.argv.includes('-v')){console.log("3.12.3");process.exit(0)}
+import{i as e}from"./chunk-EOUJ6X5H.js";function g(t,n,r,a){return{id:e(),type:t,timestamp:new Date,source:n,correlationId:a,payload:r}}var i={TestGenerated:"test-generation.TestGenerated",TestSuiteCreated:"test-generation.TestSuiteCreated",PatternLearned:"test-generation.PatternLearned",GenerationFailed:"test-generation.GenerationFailed"},s={TestRunStarted:"test-execution.TestRunStarted",TestRunCompleted:"test-execution.TestRunCompleted",FlakyTestDetected:"test-execution.FlakyTestDetected",RetryTriggered:"test-execution.RetryTriggered"},o={CoverageReportCreated:"coverage-analysis.CoverageReportCreated",CoverageGapDetected:"coverage-analysis.CoverageGapDetected",RiskZoneIdentified:"coverage-analysis.RiskZoneIdentified"},d={QualityGateEvaluated:"quality-assessment.QualityGateEvaluated",DeploymentApproved:"quality-assessment.DeploymentApproved",DeploymentBlocked:"quality-assessment.DeploymentBlocked"},l={DefectPredicted:"defect-intelligence.DefectPredicted",RootCauseIdentified:"defect-intelligence.RootCauseIdentified",RegressionRiskAnalyzed:"defect-intelligence.RegressionRiskAnalyzed"},c={KnowledgeGraphUpdated:"code-intelligence.KnowledgeGraphUpdated",ImpactAnalysisCompleted:"code-intelligence.ImpactAnalysisCompleted",SemanticSearchCompleted:"code-intelligence.SemanticSearchCompleted",C4DiagramsGenerated:"code-intelligence.C4DiagramsGenerated"},m={VulnerabilityDetected:"security-compliance.VulnerabilityDetected",ComplianceValidated:"security-compliance.ComplianceValidated",SecurityAuditCompleted:"security-compliance.SecurityAuditCompleted"},p={PatternConsolidated:"learning-optimization.PatternConsolidated",TransferCompleted:"learning-optimization.TransferCompleted",OptimizationApplied:"learning-optimization.OptimizationApplied",DreamCycleCompleted:"learning-optimization.dream.completed"},y={...i,...s,...o,...d,...l,...c,...m,...p};export{g as a,i as b,s as c,o as d,d as e,l as f,c as g,m as h,p as i};