@coldiq/mcp 0.1.19 → 0.2.5

package/tests/utils/compact-people.test.ts

@@ -0,0 +1,487 @@
+import { describe, it, expect } from 'vitest'
+import {
+  extractPeopleArray,
+  normalizePerson,
+  compactPayload,
+} from '../../src/utils/compact-people.js'
+
+// ---------------------------------------------------------------------------
+// extractPeopleArray — provider wrapper handling
+// ---------------------------------------------------------------------------
+
+describe('extractPeopleArray', () => {
+  it('leadsfactory: flattens companies_personas[i].personas (unknown[][]) into a flat list', () => {
+    const data = {
+      companies_personas: [
+        { company: 'ColdIQ', personas: [[{ contact: { name: 'Michel' } }], [{ contact: { name: 'Jane' } }]] },
+        { company: 'Folk', personas: [[{ contact: { name: 'Bob' } }]] },
+      ],
+    }
+    const arr = extractPeopleArray(data, 'leadsfactory')
+    expect(arr).toHaveLength(3)
+  })
+
+  it('leadsfactory: empty inner persona arrays produce empty result', () => {
+    expect(extractPeopleArray({ companies_personas: [{ company: 'X', personas: [[]] }] }, 'leadsfactory')).toEqual([])
+  })
+
+  it('linkupapi-search-profiles: reads data.profiles (nested wrapper)', () => {
+    const arr = extractPeopleArray({ data: { profiles: [{ name: 'A' }, { name: 'B' }] } }, 'linkupapi-search-profiles')
+    expect(arr).toHaveLength(2)
+  })
+
+  it('fullenrich / apollo / sumble / fullenrich: top-level people[]', () => {
+    expect(extractPeopleArray({ people: [{ name: 'A' }] }, 'apollo')).toHaveLength(1)
+    expect(extractPeopleArray({ people: [{ name: 'A' }] }, 'fullenrich-people-search')).toHaveLength(1)
+  })
+
+  it('pdl / companyenrich: top-level data[]', () => {
+    expect(extractPeopleArray({ data: [{ full_name: 'A' }] }, 'pdl')).toHaveLength(1)
+    expect(extractPeopleArray({ data: [{ name: 'C' }] }, 'companyenrich')).toHaveLength(1)
+  })
+
+  it('prospeo-search-person: top-level results[] (real upstream shape)', () => {
+    expect(extractPeopleArray({ results: [{ person: { full_name: 'B' } }] }, 'prospeo-search-person')).toHaveLength(1)
+  })
+
+  it('ai-ark-people: reads content[]', () => {
+    expect(extractPeopleArray({ content: [{ name: 'A' }] }, 'ai-ark-people')).toHaveLength(1)
+  })
+
+  it('findymail-search-employees: bare array passthrough', () => {
+    expect(extractPeopleArray([{ email: 'a@b.com' }], 'findymail-search-employees')).toHaveLength(1)
+  })
+
+  it('findymail-search-employees: contacts wrapper', () => {
+    expect(extractPeopleArray({ contacts: [{ email: 'a@b.com' }] }, 'findymail-search-employees')).toHaveLength(1)
+  })
+
+  it('companyenrich /people/search: items[] wrapper (real upstream shape)', () => {
+    // The CompanyEnrich /people/search endpoint returns { items: [...] } in prod;
+    // verified in src/providers/companyenrich/people-search.test.ts:45.
+    expect(extractPeopleArray({ items: [{ name: 'A' }, { name: 'B' }] }, 'companyenrich')).toHaveLength(2)
+  })
+
+  it('null/undefined/non-object returns []', () => {
+    expect(extractPeopleArray(null, 'apollo')).toEqual([])
+    expect(extractPeopleArray(undefined, 'apollo')).toEqual([])
+    expect(extractPeopleArray('whatever', 'apollo')).toEqual([])
+  })
+})
+
+// ---------------------------------------------------------------------------
+// normalizePerson — per-provider shape coverage
+// ---------------------------------------------------------------------------
+
+describe('normalizePerson — fullenrich shape', () => {
+  // Captured from a live find_people call on 2026-05-19. Trimmed to the fields
+  // we surface in compact mode + a sample of the verbose noise we drop.
+  const fullEnrichRecord = {
+    id: 'cd84e9f1-b8e7-4d7c-9b0a-1234567890ab',
+    full_name: 'Jane Smith',
+    first_name: 'Jane',
+    last_name: 'Smith',
+    location: { country: 'United States', country_code: 'US', city: 'Seattle', region: 'WA' },
+    social_profiles: {
+      professional_network: {
+        id: '12345',
+        url: 'https://www.linkedin.com/in/jane-smith',
+        handle: 'jane-smith',
+        connection_count: 500,
+      },
+    },
+    employment: {
+      current: {
+        title: 'VP of Sales',
+        seniority: 'VP',
+        description: 'Very long job description...',
+        company: {
+          id: 'co_microsoft',
+          name: 'Microsoft',
+          domain: 'microsoft.com',
+          headcount: 220000,
+          headcount_range: '10001+',
+          social_profiles: {
+            professional_network: { url: 'https://www.linkedin.com/company/microsoft' },
+          },
+          description: 'Every company has a mission...',
+          locations: { headquarters: { line1: '1 Microsoft Way' }, offices: [/* 50+ entries */] },
+          specialties: ['Business Software', 'Developer Tools', 'Quantum Computing'],
+        },
+      },
+      all: [/* heavy career history */],
+    },
+    educations: [/* ... */],
+    languages: [/* ... */],
+    skills: [/* ... */],
+  }
+
+  it('extracts compact fields and drops the heavy ones', () => {
+    const p = normalizePerson(fullEnrichRecord)!
+    expect(p).toEqual({
+      full_name: 'Jane Smith',
+      first_name: 'Jane',
+      last_name: 'Smith',
+      title: 'VP of Sales',
+      seniority: 'VP',
+      linkedin_url: 'https://www.linkedin.com/in/jane-smith',
+      company_name: 'Microsoft',
+      company_domain: 'microsoft.com',
+      company_linkedin_url: 'https://www.linkedin.com/company/microsoft',
+      company_headcount: 220000,
+      location: 'Seattle, United States',
+    })
+  })
+
+  it('stays under 500 chars per record on a trimmed FullEnrich shape', () => {
+    const compactSize = JSON.stringify(normalizePerson(fullEnrichRecord)).length
+    expect(compactSize).toBeLessThan(500)
+    // The realistic-size end-to-end check is the > 95% reduction assertion in the
+    // compactPayload suite below — this fixture is heavily trimmed and not representative.
+  })
+})
+
+describe('normalizePerson — apollo shape', () => {
+  it('reads organization.* for company info and top-level title/email', () => {
+    const p = normalizePerson({
+      id: 'apollo_123',
+      name: 'Bob Builder',
+      first_name: 'Bob',
+      last_name: 'Builder',
+      title: 'Chief Builder',
+      seniority: 'c_suite',
+      linkedin_url: 'https://linkedin.com/in/bob',
+      email: 'bob@buildco.com',
+      organization: {
+        name: 'BuildCo',
+        primary_domain: 'buildco.com',
+        linkedin_url: 'https://linkedin.com/company/buildco',
+        estimated_num_employees: 1000,
+      },
+    })!
+    expect(p).toEqual({
+      full_name: 'Bob Builder',
+      first_name: 'Bob',
+      last_name: 'Builder',
+      title: 'Chief Builder',
+      seniority: 'c_suite',
+      linkedin_url: 'https://linkedin.com/in/bob',
+      email: 'bob@buildco.com',
+      company_name: 'BuildCo',
+      company_domain: 'buildco.com',
+      company_linkedin_url: 'https://linkedin.com/company/buildco',
+      company_headcount: 1000,
+    })
+  })
+})
+
+describe('normalizePerson — pdl shape', () => {
+  it('reads job_company_* keys and job_title_levels[0]', () => {
+    const p = normalizePerson({
+      full_name: 'Charlie Chen',
+      first_name: 'Charlie',
+      last_name: 'Chen',
+      job_title: 'Director of Engineering',
+      job_title_levels: ['director'],
+      linkedin_url: 'linkedin.com/in/charlie',
+      work_email: 'charlie@biz.com',
+      job_company_name: 'BizCo',
+      job_company_website: 'biz.com',
+      job_company_linkedin_url: 'linkedin.com/company/biz',
+      job_company_size: '501-1000',
+    })!
+    expect(p).toEqual({
+      full_name: 'Charlie Chen',
+      first_name: 'Charlie',
+      last_name: 'Chen',
+      title: 'Director of Engineering',
+      seniority: 'director',
+      linkedin_url: 'linkedin.com/in/charlie',
+      email: 'charlie@biz.com',
+      company_name: 'BizCo',
+      company_domain: 'biz.com',
+      company_linkedin_url: 'linkedin.com/company/biz',
+      company_headcount: '501-1000',
+    })
+  })
+})
+
+describe('normalizePerson — leadsfactory shape', () => {
+  it('unwraps { contact: {...} } and pulls fields from inside', () => {
+    const p = normalizePerson({
+      contact: {
+        full_name: 'Dana Day',
+        first_name: 'Dana',
+        last_name: 'Day',
+        title: 'Head of Growth',
+        linkedin_url: 'https://linkedin.com/in/dana',
+        email: 'dana@growth.io',
+        company: {
+          name: 'Growth Inc',
+          domain: 'growth.io',
+          linkedin_url: 'https://linkedin.com/company/growth',
+          headcount: 50,
+        },
+      },
+    })!
+    expect(p.full_name).toBe('Dana Day')
+    expect(p.title).toBe('Head of Growth')
+    expect(p.company_name).toBe('Growth Inc')
+    expect(p.company_domain).toBe('growth.io')
+    expect(p.company_headcount).toBe(50)
+    expect(p.email).toBe('dana@growth.io')
+  })
+})
+
+describe('normalizePerson — sumble shape (real upstream fields)', () => {
+  // Verified against src/providers/sumble/route.test.ts:639-650 — Sumble /people/find
+  // returns { id, url, name, job_title, job_function, job_level, country } with NO
+  // linkedin_url field. The top-level `url` is a Sumble URL, NOT a LinkedIn URL.
+  const sumbleRecord = {
+    id: 100,
+    url: 'https://sumble.com/people/100',
+    name: 'Michel Lieben',
+    job_title: 'CEO',
+    job_function: 'sales',
+    job_level: 'executive',
+    country: 'Belgium',
+  }
+
+  it('extracts title from job_title and seniority from job_level', () => {
+    const p = normalizePerson(sumbleRecord)!
+    expect(p.title).toBe('CEO')
+    expect(p.seniority).toBe('executive')
+  })
+
+  it('does NOT surface the Sumble profile URL as linkedin_url (it is not a LinkedIn URL)', () => {
+    const p = normalizePerson(sumbleRecord)!
+    expect(p.linkedin_url).toBeUndefined()
+  })
+
+  it('accepts a real linkedin_url when the record provides one', () => {
+    const p = normalizePerson({ ...sumbleRecord, linkedin_url: 'https://www.linkedin.com/in/michel' })!
+    expect(p.linkedin_url).toBe('https://www.linkedin.com/in/michel')
+  })
+
+  it('picks up country as location', () => {
+    const p = normalizePerson(sumbleRecord)!
+    expect(p.location).toBe('Belgium')
+  })
+})
+
+describe('normalizePerson — ai-ark profile-nested shape', () => {
+  // AI-Ark /people returns { content: [{ id, profile: { full_name, ... } }] }.
+  // Verified against src/providers/ai-ark/route.test.ts:187.
+  it('unwraps record.profile so person fields are accessible to candidate paths', () => {
+    const p = normalizePerson({
+      id: 'aiark_1',
+      profile: {
+        full_name: 'Michel Lieben',
+        first_name: 'Michel',
+        last_name: 'Lieben',
+        title: 'CEO',
+        linkedin_url: 'https://www.linkedin.com/in/michel',
+      },
+    })!
+    expect(p.full_name).toBe('Michel Lieben')
+    expect(p.title).toBe('CEO')
+    expect(p.linkedin_url).toBe('https://www.linkedin.com/in/michel')
+  })
+})
+
+describe('normalizePerson — prospeo person-nested shape', () => {
+  // Prospeo /search-person returns { results: [{ person: { full_name, current_job_title, ... } }] }.
+  // Verified against the live api.prospeo.io/search-person response.
+  it('unwraps record.person and reads current_job_title as title', () => {
+    const p = normalizePerson({
+      person: {
+        person_id: 'aaaa07ad691a3a7f2ac039c6',
+        full_name: 'Catherine Powlett',
+        first_name: 'Catherine',
+        last_name: 'Powlett',
+        current_job_title: 'Marketing Manager',
+        linkedin_url: 'https://www.linkedin.com/in/catherinepowlett',
+      },
+    })!
+    expect(p.full_name).toBe('Catherine Powlett')
+    expect(p.first_name).toBe('Catherine')
+    expect(p.last_name).toBe('Powlett')
+    expect(p.title).toBe('Marketing Manager')
+    expect(p.linkedin_url).toBe('https://www.linkedin.com/in/catherinepowlett')
+  })
+})
+
+describe('normalizePerson — leadsfactory persona-level company sibling', () => {
+  // LF persona items have shape { contact, company, persona_index, persona_job_title }.
+  // Verified against src/providers/leadsfactory/schema.ts:134-141. The company sibling
+  // must remain accessible to company_* candidate paths after unwrapping contact.
+  it('reads company.* from the persona sibling, not just contact-nested fields', () => {
+    const p = normalizePerson({
+      contact_found: true,
+      persona_index: 0,
+      persona_job_title: 'CEO',
+      contact: {
+        full_name: 'Michel Lieben',
+        title: 'CEO',
+        linkedin_url: 'https://www.linkedin.com/in/michel',
+      },
+      company: {
+        name: 'ColdIQ',
+        domain: 'coldiq.com',
+        linkedin_url: 'https://www.linkedin.com/company/coldiq',
+        headcount: 12,
+      },
+    })!
+    expect(p.full_name).toBe('Michel Lieben')
+    expect(p.title).toBe('CEO')
+    expect(p.company_name).toBe('ColdIQ')
+    expect(p.company_domain).toBe('coldiq.com')
+    expect(p.company_linkedin_url).toBe('https://www.linkedin.com/company/coldiq')
+    expect(p.company_headcount).toBe(12)
+  })
+})
+
+describe('normalizePerson — linkedin URL strict validation', () => {
+  it('rejects non-LinkedIn URLs in url/profile_url/linkedin fallbacks', () => {
+    const p = normalizePerson({
+      name: 'X',
+      profile_url: 'https://example.com/profile/x',
+      url: 'https://sumble.com/people/100',
+    })!
+    expect(p.linkedin_url).toBeUndefined()
+  })
+
+  it('accepts LinkedIn URLs at fallback paths', () => {
+    const p = normalizePerson({
+      name: 'X',
+      profile_url: 'https://www.linkedin.com/in/x',
+    })!
+    expect(p.linkedin_url).toBe('https://www.linkedin.com/in/x')
+  })
+
+  it('accepts PDL canonical short form (no protocol, no www)', () => {
+    const p = normalizePerson({ full_name: 'X', linkedin_url: 'linkedin.com/in/x' })!
+    expect(p.linkedin_url).toBe('linkedin.com/in/x')
+  })
+})
+
+describe('normalizePerson — edge cases', () => {
+  it('composes full_name from first + last when full_name missing', () => {
+    const p = normalizePerson({ first_name: 'Anna', last_name: 'Apple' })!
+    expect(p.full_name).toBe('Anna Apple')
+  })
+
+  it('drops fully-empty records', () => {
+    expect(normalizePerson({})).toBeNull()
+    expect(normalizePerson(null)).toBeNull()
+    expect(normalizePerson('string')).toBeNull()
+  })
+
+  it('drops empty-string / null candidates and falls through to the next path', () => {
+    const p = normalizePerson({
+      title: '',
+      job_title: 'Fallback Title',
+    })!
+    expect(p.title).toBe('Fallback Title')
+  })
+
+  it('omits fields that resolved to nothing (no undefined leaks into output)', () => {
+    const p = normalizePerson({ full_name: 'Solo Person' })!
+    expect(Object.keys(p)).toEqual(['full_name'])
+  })
+
+  it('handles array index in path (job_title_levels.0)', () => {
+    const p = normalizePerson({ name: 'X', job_title_levels: ['vp', 'manager'] })!
+    expect(p.seniority).toBe('vp')
+  })
+})
+
+// ---------------------------------------------------------------------------
+// compactPayload — gap_fill folding + total passthrough
+// ---------------------------------------------------------------------------
+
+describe('compactPayload', () => {
+  it('folds gap_fill (apollo) results into the same people[] and reports gap_fill_provider', () => {
+    const data = {
+      companies_personas: [
+        { company: 'ColdIQ', personas: [[{ contact: { full_name: 'Michel L', title: 'CEO' } }]] },
+      ],
+      gap_fill: {
+        provider: 'apollo',
+        domains: ['nohit.com'],
+        people: [
+          {
+            name: 'Apollo Adam',
+            title: 'CTO',
+            organization: { name: 'NoHit', primary_domain: 'nohit.com' },
+          },
+        ],
+      },
+    }
+    const out = compactPayload(data, 'leadsfactory')
+    expect(out.people).toHaveLength(2)
+    expect(out.people[0].full_name).toBe('Michel L')
+    expect(out.people[1].full_name).toBe('Apollo Adam')
+    expect(out.people[1].company_domain).toBe('nohit.com')
+    expect(out.gap_fill_provider).toBe('apollo')
+  })
+
+  it('passes metadata.total through (fullenrich pagination)', () => {
+    const data = {
+      people: [{ full_name: 'A' }],
+      metadata: { total: 42, credits: 1, offset: 0 },
+    }
+    const out = compactPayload(data, 'fullenrich-people-search')
+    expect(out.total).toBe(42)
+  })
+
+  it('omits gap_fill_provider when there is no gap_fill block', () => {
+    const out = compactPayload({ people: [{ full_name: 'A' }] }, 'apollo')
+    expect(out.gap_fill_provider).toBeUndefined()
+  })
+
+  it('returns empty people[] when data is null or empty', () => {
+    expect(compactPayload(null, 'apollo').people).toEqual([])
+    expect(compactPayload({}, 'apollo').people).toEqual([])
+  })
+
+  it('end-to-end size check: a realistic fullenrich-like payload shrinks dramatically', () => {
+    const heavy = {
+      people: Array.from({ length: 5 }).map((_, i) => ({
+        id: `person_${i}`,
+        full_name: `Person ${i}`,
+        first_name: `Person`,
+        last_name: String(i),
+        social_profiles: { professional_network: { url: `https://linkedin.com/in/person${i}` } },
+        location: { country: 'United States', city: 'Seattle' },
+        employment: {
+          current: {
+            title: 'VP of Sales',
+            seniority: 'VP',
+            description: 'x'.repeat(1500),
+            company: {
+              name: 'Microsoft',
+              domain: 'microsoft.com',
+              headcount: 220000,
+              description: 'x'.repeat(800),
+              locations: { headquarters: { line1: '1 Microsoft Way' }, offices: Array.from({ length: 40 }).map(() => ({ line1: 'Office', line2: 'somewhere' })) },
+              specialties: Array.from({ length: 25 }).map((_, j) => `Specialty ${j}`),
+              social_profiles: { professional_network: { url: 'https://linkedin.com/company/microsoft' } },
+            },
+          },
+          all: Array.from({ length: 10 }).map(() => ({ title: 'Past role', description: 'y'.repeat(1000) })),
+        },
+        skills: Array.from({ length: 30 }).map((_, j) => `skill_${j}`),
+        languages: Array.from({ length: 5 }).map(() => ({ language: 'English', proficiency: 'Native' })),
+        educations: Array.from({ length: 4 }).map(() => ({ school_name: 'Some University' })),
+      })),
+      metadata: { total: 5 },
+    }
+    const rawSize = JSON.stringify(heavy).length
+    const compact = compactPayload(heavy, 'fullenrich-people-search')
+    const compactSize = JSON.stringify(compact).length
+    expect(compactSize).toBeLessThan(rawSize / 20) // > 95% reduction
+    expect(compact.people).toHaveLength(5)
+    expect(compact.total).toBe(5)
+  })
+})