generate_verification_plan

Instructions

Implementation Reference

• src/engine/VerificationPlanningEngine.ts:4-53 (handler)

  Core business logic: iterates risk reports and generates tiered verification targets (Structural, Behavioral, Adversarial) with directives and execution recommendations.

  export class VerificationPlanningEngine {
      
      public generatePlan(riskReports: RiskReport[]): VerificationPlan {
          const targets: VerificationTarget[] = [];
          const executionRecommendations = new Set<string>();
  
          riskReports.forEach(report => {
              const { score, nodeId } = report;
  
              // Tier 1: Always do basic structural checks for impacted nodes
              targets.push({
                  nodeId,
                  tier: VerificationTier.Structural,
                  directive: `Run linting, schema validation, and type-checks for ${nodeId}`,
                  priority: 'Low'
              });
  
              // Tier 2: If integration count is noticeable or moderate risk
              if (score.integrationCount > 2 || score.overallRisk > 30) {
                  targets.push({
                      nodeId,
                      tier: VerificationTier.Behavioral,
                      directive: `Validate workflow correctness and integration contracts pointing to/from ${nodeId}`,
                      priority: 'Medium'
                  });
                  executionRecommendations.add('Run integration tests simulating dependent service data.');
              }
  
              // Tier 3: High Blast Radius or Runtime Criticality dictates adversarial logic
              if (score.blastRadius > 50 || score.runtimeCriticality >= 80) {
                  targets.push({
                      nodeId,
                      tier: VerificationTier.Adversarial,
                      directive: `Check concurrency, malformed state handling, and race conditions for ${nodeId}`,
                      priority: 'High'
                  });
                  executionRecommendations.add('Conduct adversarial testing mapping race conditions against active DB constraints.');
                  executionRecommendations.add('Trigger manual QA or specialized Autonomous QA Agent for critical paths.');
              }
          });
  
          // Default global recommendations
          executionRecommendations.add('Delegate dynamic tests to CI execution pipelines.');
  
          return {
              targets,
              executionRecommendations: Array.from(executionRecommendations)
          };
      }
  }

• src/models/VerificationModels.ts:1-17 (schema)

  Type definitions for VerificationTier enum, VerificationTarget interface, and VerificationPlan interface.

  export enum VerificationTier {
      Structural = 'Tier 1 - Structural Verification',
      Behavioral = 'Tier 2 - Behavioral Verification',
      Adversarial = 'Tier 3 - Adversarial Verification'
  }
  
  export interface VerificationTarget {
      nodeId: string;
      tier: VerificationTier;
      directive: string;
      priority: 'High' | 'Medium' | 'Low';
  }
  
  export interface VerificationPlan {
      targets: VerificationTarget[];
      executionRecommendations: string[];
  }