What problem does it solve? Enriching every contact found in a people search wastes enrichment credits on poor-fit leads. This Skill filters and ranks contacts from people-search output before enrichment, rejecting non-ICP contacts and prioritizing the best matches so only qualified contacts consume paid enrichment credits. ## Core Features & Use Cases - Multi-Tier ICP Classification: Scores contacts on job tier (1-6), industry tier (0-5), location tier (A-D), company size, and ICP keyword matches in company specialities, with configurable hard-reject thresholds. - Smart Edge-Case Handling: Applies C-level promotion at small companies, demotes assistant/support titles even when they contain decision-maker keywords, and falls back gracefully when company data is missing. - Configurable Ranking: Sorts passed contacts by a client-defined composite sort key stored in context/icp.md, with dedup by LinkedIn URL and best-N-per-company selection for small batches. - Use Case: After a people-search returns 500 contacts, run this filter to reject interns and out-of-region roles, rank the rest by job and industry fit, and pass only the top 150 to people-enrichment — cutting enrichment spend significantly. ## Quick Start Filter and rank the contacts in csv/intermediate/contacts_found.csv using my ICP configuration and write the qualified results to contacts_filtered.csv.
