Lead Enrichment Scanner

Automated B2B lead generation and enrichment. Input: company names or domains. Output: enriched contact records with decision-maker names, emails, LinkedIn p...

When to use this skill

Activate when the user wants to:

• Look up basic info on a company from its public website
• Generate email address patterns for a company
• Get a company's description and social links from their homepage
• Run enrichment in dry-run mode to preview output

v1.0 limitations: This is a basic scraper + optional LLM enrichment. Decision-maker names, LinkedIn profiles, verified contacts, CRM import, and outreach drafts are NOT implemented (v1.1).

Prerequisites

• MINIMAX_API_KEY (optional — only needed for LLM-based description extraction)
• OpenClaw gateway running

Workflow

1. Prepare input

Create a text file with company names (one per line):

bash

cat > companies.txt <<EOF
Stripe
Notion
Linear
EOF

2. Run enrichment

bash

clawhub run lead-enrichment-scanner enrich \
  --input companies.txt \
  --output leads.csv

What v1.0 actually extracts:

• Best-guess domain from company name
• Homepage description (scraped via HTTPS)
• Email address patterns (guessed, not verified)
• Social links found on homepage
• Industry/company size via LLM (optional, MINIMAX_API_KEY required)

3. Dry-run mode

bash

clawhub run lead-enrichment-scanner enrich \
  --input companies.txt \
  --output leads.csv \
  --dry-run

4. Configure respect for robots.txt

bash

clawhub run lead-enrichment-scanner configure \
  --respect-robots-txt true \
  --rate-limit-seconds 5

What v1.0 actually does

• ✅ Domain guessing from company name
• ✅ Public homepage scrape (HTTPS only, respects robots.txt)
• ✅ Email pattern generation (guessed, not verified)
• ✅ LLM metadata extraction (optional, MINIMAX_API_KEY required)
• ✅ SSRF protection (blocks private IPs, HTTPS-only)
• ❌ Decision-maker names — NOT implemented (v1.1)
• ❌ LinkedIn profiles — NOT implemented (v1.1)
• ❌ Verified contact emails — NOT implemented (v1.1)
• ❌ CRM import — NOT implemented (v1.1)
• ❌ Outreach draft generation — NOT implemented (v1.1)
• ❌ Bulk parallel processing (--bulk) — NOT implemented (v1.1)

Output fields

• name — company name as provided
• domain — guessed domain
• description — from homepage scrape or LLM (if key set)
• email_patterns — guessed email formats (e.g. {firstname}@domain.com)
• social_links — LinkedIn/Twitter/GitHub links found on homepage

Important: Email patterns are guesses, not verified deliverable addresses. Do not send outreach without verifying addresses first.

Compliance responsibility

• robots.txt is respected by default
• Rate-limited to 5s between requests
• Scrapes public homepage content only
• You are responsible for GDPR/CAN-SPAM compliance when using contact data
• Do not provide LinkedIn session cookies — they are not used in v1.0

Cost expectations

• Web scraping: free (public data, HTTPS)
• LLM extraction: ~$0.005 per company (only when MINIMAX_API_KEY is set)
• OpenClaw: free, self-hosted

Troubleshooting

Empty description for a company?

• The homepage may block scraping — try a different domain format

Email patterns look wrong?

• Email patterns are guesses, not verified addresses. Always verify before outreach.

Want LinkedIn enrichment, CRM import, or outreach drafts?

• Planned for v1.1. Not implemented in v1.0.