Contact Enricher

Given a list of company domains or URLs, crawls a handful of high-signal pages per site (home, contact, about, team, footer links, JSON-LD), extracts emails, phones, socials, address, company name and tech hints, infers the company's email pattern (e.g. {first}.{last}), optionally verifies emails via MX lookup + catch-all detection, and emits one normalised record per domain.

Input

Field	Type	Default	Description
domains	string[]	(required)	Domains or URLs to enrich, e.g. "acme.com" or "https://www.acme.com". urls is accepted as an alias.
maxPagesPerSite	integer	8	Max pages crawled per domain (home + up to maxPagesPerSite - 1 internal links).
verifyEmails	boolean	true	Run MX + catch-all checks on every extracted/guessed email.
wantPersonNames	boolean	false	Include named individuals found on the site and generate up to 10 pattern-guessed personal emails. Off by default — business contact data only, per policy.
verifierProvider	string	(none)	Optional third-party verifier API name (see src/verify.py adapter).
verifierApiKey	string	(none)	API key for the third-party verifier, if used.
maxConcurrency	integer	5	Domains processed in parallel.

Output (one dataset row per domain)

{
  "domain": "acme.com",
  "company_name": "Acme Corp",
  "emails": [
    {
      "value": "info@acme.com",
      "type": "generic",
      "source": "observed",
      "off_domain": false,
      "mx_ok": true,
      "smtp_status": "unknown",
      "confidence": 0.9
    },
    {
      "value": "jane.doe@acme.com",
      "type": "named",
      "source": "observed",
      "off_domain": false,
      "mx_ok": true,
      "smtp_status": "unknown",
      "confidence": 0.9
    },
    {
      "value": "john.smith@acme.com",
      "type": "named",
      "source": "pattern",
      "off_domain": false,
      "mx_ok": true,
      "smtp_status": "unknown",
      "confidence": 0.6
    }
  ],
  "email_pattern": "{first}.{last}",
  "phones": ["+15551234567"],
  "socials": { "linkedin": "https://linkedin.com/company/acme" },
  "address": "123 Main St, Springfield, IL, 62704, US",
  "tech_hints": { "cms": "wordpress", "analytics": "google-analytics" },
  "pages_crawled": 5,
  "source_urls": [
    "https://acme.com/",
    "https://acme.com/contact",
    "https://acme.com/about",
    "https://acme.com/team"
  ],
  "fetched_at": "2026-08-29T12:00:00+00:00"
}

A failed domain produces {"domain": "...", "error": "..."} instead, and is never charged.

When wantPersonNames is true, a person_names array is also included.

Pricing events (pay-per-event)

Event	Suggested price	When
domain-enriched	$0.002	Once per successfully processed domain (one dataset row).
email-verified	$0.004	Per email that went through verify_email (only when verifyEmails is true).

Failed domains are never charged.

MCP

Exposed as an MCP tool enrich_company(domain), which maps to running this Actor with domains: [domain] and default options, returning the single resulting record.

Limitations

• JS-only sites: the crawler is a plain HTTP client (httpx) with no headless browser — content injected client-side by JavaScript frameworks won't be seen.
• SMTP probing is often blocked: outbound port 25 is frequently blocked from datacenter/cloud IPs, so catch-all detection and per-mailbox checks regularly come back inconclusive. In that case smtp_status is "unknown" (still usable — mx_ok and confidence are still populated from the MX lookup).
• Only business contact data is emitted by default; named individuals and pattern-guessed personal emails require the explicit wantPersonNames flag.
• Email-pattern inference and address/company-name selection are heuristic, not guaranteed correct — always treat guessed (source: "pattern") emails as lower confidence than observed ones.