icp-scoringlisted

# Skill · icp-scoring ## What this does Given a list of accounts (domains), produces a tier-ranked output with disqualifier flags and per-account first-action. Different from traditional ICP scoring because the gates are AI-native (Maturity Stage, Data Residency, Spend Trajectory) — see `context/icp-definition.md` Section 6. ## How to invoke ``` score these accounts: [list of domains] ``` Or: ``` score the accounts in [outputs/file.md] ``` ## Steps Claude takes 1. **Read** `context/icp-definition.md` Sections 1, 2, 3, 6 (qualifier framework + anti-ICP). 2. **For each account**, run gates in this order: **Gate 1 · Hard qualifiers** (any fail → disqualify): - LLM Maturity Stage in target range? - Data Residency posture compatible? - Decision authority for AI tooling exists in target role? **Gate 2 · Disqualifiers** (any hit → reject): - Anti-ICP pattern triggered? - Customer's primary moat is being a competitor's wrapper? - Customer asks "swap the model" in first conversation? **Gate 3 · Soft scoring** (sum points): - Industry in target list: +2 - Champion identified (specific name): +2 - Model spend trajectory growing 30%+: +3 - Already uses a competitor we displace well: +2 - Already uses a complementary tool we integrate with: +1 - Has shipped public MCP server / Agent Skill (Constitution #6): +2 - Founder publicly active on AI-native topics: +1 3. **Assign tier**: | Tier | Criteria | Action | |---|---|---|