codecrucible-synth
Version:
Production-Ready AI Development Platform with Multi-Voice Synthesis, Smithery MCP Integration, Enterprise Security, and Zero-Timeout Reliability
120 lines • 3.02 kB
TypeScript
export interface BenchmarkResult {
metric: string;
averageTime: number;
speedupFactor?: number;
documentedClaim: string;
meetsDocumentedMetric: boolean;
details: {
iterations: number;
times: number[];
baseline?: number;
target?: number;
tasks?: TaskResult[];
};
}
export interface TaskResult {
name: string;
ollamaTime: number;
hybridTime: number;
improvement: number;
accuracy: number;
}
export interface ValidationReport {
validatedClaims: BenchmarkResult[];
invalidClaims: BenchmarkResult[];
recommendations: string[];
overallScore: number;
timestamp: number;
environment: SystemEnvironment;
}
export interface SystemEnvironment {
platform: string;
nodeVersion: string;
memory: number;
cpu: string;
gpu?: string;
timestamp: number;
}
export interface ResourceMetrics {
memoryUsage: number;
vramUsage?: number;
cpuUsage: number;
timestamp: number;
}
/**
* Performance Validation Framework matching documented benchmarks
* Validates all claims from Performance-Benchmarks.md
*/
export declare class PerformanceValidator {
private hybridClient?;
private ollamaClient?;
private lmStudioClient?;
private metrics;
private resourceMetrics;
constructor();
/**
* Validate all documented performance claims
*/
validateDocumentationClaims(): Promise<ValidationReport>;
/**
* Benchmark template generation - targeting 19x improvement
*/
private benchmarkTemplateGeneration;
/**
* Benchmark code formatting - targeting 25x improvement
*/
private benchmarkCodeFormatting;
/**
* Benchmark first token latency - targeting sub-200ms
*/
private benchmarkFirstTokenLatency;
/**
* Benchmark simple edits - targeting 14x improvement
*/
private benchmarkSimpleEdits;
/**
* Benchmark boilerplate code generation - targeting 16x improvement
*/
private benchmarkBoilerplateCode;
/**
* Benchmark complex analysis - should maintain quality
*/
private benchmarkComplexAnalysis;
/**
* Benchmark task routing accuracy - targeting 94.5%
*/
private benchmarkTaskRouting;
/**
* Benchmark resource utilization
*/
private benchmarkResourceUtilization;
/**
* Measure execution time of a task
*/
private measureTaskTime;
/**
* Extract task type from prompt
*/
private extractTaskType;
/**
* Initialize LLM clients for testing
*/
private initializeClients;
/**
* Get system environment information
*/
private getSystemEnvironment;
/**
* Generate recommendations based on benchmark results
*/
private generateRecommendations;
/**
* Save benchmark report to file
*/
private saveReport;
/**
* Cleanup resources
*/
dispose(): Promise<void>;
}
//# sourceMappingURL=performance-validator.d.ts.map