Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Two-stage contact filtering and scoring for personal CRM workflows. A fast keyword pass runs first, then an AI-powered relevance score for deeper analysis. The system records your feedback and adjusts its rankings over time.
The two-stage approach keeps costs low by running AI scoring only on contacts that pass the initial filter, rather than sending every record to the LLM.
When to use it
- Prioritizing follow-ups after a conference
- Filtering a large imported contact list down to warm leads
- Scoring a new contact before a sales call
- Finding who to reach out to this week from a stale CRM
- Reviewing which contacts match a specific campaign context
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: filter my contacts and show me the top 10 leads
- 1Run stage 1 keyword filter against contacts.json to discard low-signal entries
- 2Pass remaining contacts through stage 2 AI relevance scoring via Anthropic API
- 3Rank contacts by score and return the top 10 as JSON
- 4Update learning.json with session results for future preference tuning
JSON list of top 10 scored contacts with relevance scores and ranking rationale
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.