icp-scorer - SKILL.md Agent Skill

name: icp-scorer description: Score leads against your Ideal Customer Profile. Analyzes title, company, industry fit and assigns A/B/C/D tiers for prioritization.

ICP Scorer

Score leads against your Ideal Customer Profile to prioritize outreach.

What It Does

Takes enriched leads from lead-enricher
Analyzes each lead's title, company, industry
Scores against configurable ICP criteria
Assigns tiers: A (80+), B (60-79), C (40-59), D (<40)

Invocation

# Score leads with default ICP
./scripts/score.sh leads.json > scored.json

# Score with custom ICP config
ICP_CONFIG=my-icp.json ./scripts/score.sh leads.json > scored.json

# Filter to A/B tier only
./scripts/score.sh leads.json | jq '.leads | map(select(.tier == "A" or .tier == "B"))'

Setup

Environment Variables

# For AI-powered scoring (optional but recommended)
export ANTHROPIC_API_KEY="sk-ant-..."
# OR
export OPENAI_API_KEY="sk-..."

ICP Configuration

Create icp-config.json:

{
  "target_titles": [
    "VP Marketing",
    "Head of Marketing", 
    "CMO",
    "Director of Marketing",
    "Head of Growth",
    "Founder",
    "CEO",
    "Agency Owner"
  ],
  "target_industries": [
    "SaaS",
    "E-commerce",
    "DTC",
    "Marketing Agency",
    "Technology"
  ],
  "company_size": {
    "min": 10,
    "max": 500,
    "ideal": "50-200"
  },
  "keywords": [
    "AI", "automation", "marketing", "growth", "scale"
  ],
  "exclude_titles": [
    "Student",
    "Intern",
    "Looking for",
    "Open to work"
  ]
}

Scoring Rubric

Category	Points	Criteria
Title Match	0-25	Exact match = 25, Close = 20, Related = 15, Adjacent = 10
Company Size	0-20	Perfect fit = 20, In range = 15, Close = 10
Industry	0-20	Exact match = 20, Adjacent = 15, Related = 10
Keywords	0-15	Multiple matches = 15, Some = 10, Few = 5
Engagement	0-10	From intent_score (comment > reaction)
Signals	0-10	Hiring, funding, growth mentions

Total: 100 points

Tier Assignment

Tier	Score	Action
A	80-100	Priority outreach, heavy personalization
B	60-79	Include in sequence, moderate personalization
C	40-59	Lower priority, template outreach
D	0-39	Skip or save for later

Input Format

Expects output from lead-enricher:

{
  "leads": [
    {
      "name": "Jane Doe",
      "headline": "VP Marketing at Acme Corp | AI Enthusiast",
      "company": "Acme Corp",
      "company_size": "51-200",
      "company_industry": "SaaS",
      "linkedin_url": "https://linkedin.com/in/janedoe",
      "intent_score": 10,
      "engagement_text": "Great insights on AI automation!"
    }
  ]
}

Output Format

{
  "meta": {
    "total_leads": 50,
    "tier_a": 8,
    "tier_b": 15,
    "tier_c": 12,
    "tier_d": 15,
    "scored_at": "2026-02-14T15:00:00Z"
  },
  "leads": [
    {
      "name": "Jane Doe",
      "headline": "VP Marketing at Acme Corp | AI Enthusiast",
      "company": "Acme Corp",
      "linkedin_url": "https://linkedin.com/in/janedoe",
      "icp_score": 85,
      "tier": "A",
      "score_breakdown": {
        "title_match": 25,
        "company_size": 20,
        "industry": 20,
        "keywords": 10,
        "engagement": 10,
        "signals": 0
      },
      "score_reason": "Exact title match (VP Marketing), ideal company size, SaaS industry, AI keywords in headline"
    }
  ]
}

Scoring Modes

Mode 1: Rule-Based (Fast, Free)

Uses keyword matching and rules. Good for high volume.

SCORING_MODE=rules ./scripts/score.sh leads.json

Mode 2: AI-Powered (Better, Costs Tokens)

Uses Claude/GPT to analyze fit. Better for nuanced scoring.

SCORING_MODE=ai ./scripts/score.sh leads.json

Usage in First 1000 Kit

# Full pipeline
./linkedin-miner/scripts/mine.sh "AI marketing" 20 > raw.json
./lead-enricher/scripts/profile-enrich.sh raw.json | \
  ./lead-enricher/scripts/enrich.sh > enriched.json
./icp-scorer/scripts/score.sh enriched.json > scored.json

# Get A-tier leads only
cat scored.json | jq '.leads | map(select(.tier == "A"))'

Example ICP Configs

For Agency Services

{
  "target_titles": ["Founder", "CEO", "Agency Owner", "Managing Director"],
  "target_industries": ["Marketing Agency", "Creative Agency", "Digital Agency"],
  "company_size": {"min": 5, "max": 50}
}

For SaaS Product

{
  "target_titles": ["VP Marketing", "Head of Growth", "CMO", "Director Marketing"],
  "target_industries": ["SaaS", "Technology", "Software"],
  "company_size": {"min": 50, "max": 500}
}

For E-commerce/DTC

{
  "target_titles": ["Founder", "CEO", "Head of Marketing", "E-commerce Manager"],
  "target_industries": ["E-commerce", "DTC", "Retail", "Consumer Goods"],
  "company_size": {"min": 10, "max": 200}
}