LeadIntel OSINT Skill

Purpose

Validate B2B leads as textile finishing mill prospects using dual evidence (K1 external + K2 internal) and assign actionable tier classification.

Core Rules

1. Mill Definition

A "Prospect Mill" must:

• •Operate finishing/dyeing/processing equipment (NOT just trade/distribute)
• •Serve the apparel or technical textile industry
• •Have operational capacity (not startup-stage only)

Machines in scope: Stenter frames, tenter frames, ramöz, dyeing vats, mercerizing lines, coating applicators, heat-setting ovens

2. Evidence Standards (K1 + K2)

Tier-1 Golden: BOTH K1 (≥1) AND K2 (≥1)
Tier-2 Promising: K1 only OR K2 only
Tier-3 Research: No evidence yet, or contradictory signals

3. Negative Filters (Auto-Reject)

• •Company name contains: Brand, Retail, Trading, Distribution, Group, Association, Event
• •Website is generic listing (*.org.br, global-trace-base.org, linkedin.com only)
• •Contact role is sales/marketing for unrelated product
• •No operational mill found after 20-min research

4. Classification Logic

def classify_lead(k1_count, k2_count, negative_signal, oem_brand):
    if negative_signal:
        return "❌ REJECT"
    if k1_count >= 1 and k2_count >= 1:
        return "✅ TIER-1 GOLDEN"
    if k1_count >= 1 or k2_count >= 1:
        return "⚠️ TIER-2 PROMISING"
    return "🔍 TIER-3 RESEARCH"

Skill Application Examples

Example 1: Clear Fit

Input: Acme Textile Mills, Turkey, no website provided

Process:

1. •Search: "Acme Textile Mills Turkey dyeing" → Find acme-textile.com.tr
2. •Check Monforts reference list → Found "Acme Mills"
3. •Visit acme-textile.com.tr/capabilities → "Monforts stenter, heat-setting, 50 tons/day"
4. •Search LinkedIn → Job post: "Stenter Operator – Acme Mills (posted 2 months ago)"

Output:

{
  "classification": "✅ TIER-1 GOLDEN",
  "k1_evidence": ["Monforts reference list", "LinkedIn job posting"],
  "k2_evidence": ["Website /capabilities page"],
  "hs_code": "8451.90",
  "sales_angle": "Monforts chain parts + spindle nuts for 50-ton stenter",
  "next_action": "Prioritize outreach; reference Monforts customer base"
}

Example 2: Uncertain Fit

Input: TechFab Solutions, Pakistan

Process:

1. •Website: techfab.pk → "Textile solutions provider"
2. •No stenter/finishing keywords visible
3. •LinkedIn: "B2B platform for textile buyers and sellers"
4. •NOT found in OEM reference lists
5. •No job postings for stenter operators

Output:

{
  "classification": "⚠️ TIER-2 PROMISING (Low)",
  "reason": "Website vague; no K1 evidence yet",
  "next_action": "Request more details: Do you operate finishing equipment? What brands?",
  "confidence": "MEDIUM"
}

Example 3: Clear Reject

Input: Istanbul Fashion Group

Process:

1. •Website: "Fashion brand showcasing Turkish textiles"
2. •No production capability mentioned
3. •Retail/distribution focus
4. •NOT in OEM reference lists

Output:

{
  "classification": "❌ REJECT",
  "reason": "Retail brand, not mill operator",
  "confidence": "HIGH"
}

When Applying This Skill

Use this skill when:

• •✅ Validating a CSV of new leads from fairs/trade data
• •✅ Double-checking a lead's "Tier" before CRM entry
• •✅ Scoring HS code fit (8451.90 vs fallback)
• •✅ Generating "Why this customer?" narratives for sales

Do NOT use if:

• •❌ Lead has clear legal/compliance issues (IP disputes, sanctions)
• •❌ User is asking about proprietary competitor data
• •❌ Scraping LinkedIn (ToS violation)

Output Checklist

• • Company name + country confirmed
• • Website verified (or flagged as missing)
• • K1 evidence cited with URL
• • K2 evidence cited with excerpt + keyword list
• • Tier classification assigned (✅ / ⚠️ / ❌)
• • HS code prediction (8451.90 or fallback reasoning)
• • Next sales action stated
• • Confidence level (HIGH/MEDIUM/LOW)

Standard: LeadIntel Pro V9 Sniper (K1+K2 Dual Evidence)
Skill Version: 1.0
Last Updated: 2026-02-05