/** * Threat Lab - Deep Research via modelab * * When a finding is flagged (static/deps/sim), this module can be invoked * to run the finding through modelab's iterative multi-model research * pipeline for: * * 1. Root cause analysis - multi-model perspective on the vulnerability * 2. Impact assessment - how severe is this really? * 3. Patch generation - concrete, validated fix proposals * 4. Exploit scenario - how would an attacker actually use this? * * Usage: threat-lab scan . --deep * * Requires BANKR_API_KEY (for the underlying LLM calls). * modelab routes requests through the Bankr LLM gateway. */ export interface DeepResearchFinding { category: 'static' | 'deps' | 'sim'; severity: 'critical' | 'high' | 'medium' | 'low' | 'informational'; title: string; description: string; packageName?: string; contractFile?: string; evidence?: string; } export interface DeepResearchResult { finding: DeepResearchFinding; rootCause: string; impactAssessment: string; exploitScenario: string; proposedFix: PatchFix | null; modelsUsed: string[]; totalCost: number; durationMs: number; } export interface PatchFix { title: string; description: string; code: string; diff: string; confidence: number; validation: string; references: string[]; } export interface DeepResearchOptions { finding: DeepResearchFinding; contractCode?: string; models?: string[]; } export declare function runDeepResearch(options: DeepResearchOptions): Promise; export interface DeepResearchBatchOptions { findings: DeepResearchFinding[]; contractCode?: string; models?: string[]; maxFindings?: number; } export declare function runDeepResearchBatch(options: DeepResearchBatchOptions): Promise; export declare function formatDeepResearchReport(results: DeepResearchResult[]): string; //# sourceMappingURL=deepResearch.d.ts.map