RelayPlane

/** * Lead Enrichment Skill * * Enrich lead data with company info, social profiles, and qualification scores. * * ## Context Efficiency * Without this skill: ~10k tokens (lookup + analysis + formatting all in context) * With this skill: ~1k tokens (only enriched lead data in context) * **Context reduction: 90%** * * ## Models Used * - Claude 3.5 Haiku: Analysis (fast, good reasoning) * - GPT-4o-mini: Formatting (fast, cheap) */ export const leadEnrichmentConfig = { name: 'lead-enrichment', description: 'Enrich lead data with company info, social profiles, and qualification scores.', category: 'integration', contextReduction: '90%', steps: [ { name: 'lookup', // This would use an MCP tool in real implementation model: 'openai:gpt-4o-mini', prompt: `Given this lead email, extract what we know: Email: {{input.email}} Name: {{input.name}} Infer and provide: 1. Likely company name and domain 2. Probable role/title 3. Industry 4. Company size estimate This is for lead qualification purposes.`, }, { name: 'analyze', model: 'anthropic:claude-3-5-haiku-20241022', depends: ['lookup'], prompt: `Score this lead based on the enriched data: Lead data: {{steps.lookup.output}} Provide: 1. Lead score (1-100) 2. Qualification tier (Hot/Warm/Cold) 3. Key signals (positive and negative) 4. Recommended next action 5. Best contact channel Be concise but thorough.`, }, { name: 'format', model: 'openai:gpt-4o-mini', depends: ['analyze'], prompt: `Format this lead analysis for CRM import: Analysis: {{steps.analyze.output}} Original lead: {{input.email}}, {{input.name}} Create a structured JSON object suitable for CRM import with: - contact info - company info - lead score - qualification tier - notes - recommended actions`, schema: { type: 'object', properties: { contact: { type: 'object', properties: { email: { type: 'string' }, name: { type: 'string' }, title: { type: 'string' }, }, }, company: { type: 'object', properties: { name: { type: 'string' }, domain: { type: 'string' }, industry: { type: 'string' }, size: { type: 'string' }, }, }, qualification: { type: 'object', properties: { score: { type: 'number' }, tier: { type: 'string' }, signals: { type: 'array', items: { type: 'string' } }, }, }, nextAction: { type: 'string' }, notes: { type: 'string' }, }, }, }, ], estimatedCost: '$0.01-0.03 per lead', }; /** * Get the workflow configuration for lead enrichment */ export function getLeadEnrichmentWorkflow(input: { email: string; name?: string }) { return { name: leadEnrichmentConfig.name, steps: leadEnrichmentConfig.steps, input: { email: input.email, name: input.name || '', }, }; }