What problem does it solve? Support teams struggle to understand what their Case volume actually contains and which requests can be safely automated. This Skill turns raw Salesforce support Cases into an evidence-based demand analysis and a concrete LangGraph agent design grounded in that data. ## Core Features & Use Cases - Case Clustering: Pulls Cases via read-only SOQL and clusters them by issue category and resolution pattern, delegating large volumes to a sub-agent. - Automation Quantification: Classifies demand into self_service, agent_assisted, and human_required handling splits, reported as scenario coverage rather than deflection KPIs. - Agent Design Output: Produces a triage-first LangGraph design (triage, retrieve, respond, escalate, resolve) with narrow tools and human-in-the-loop escalation for fraud and legal cases. - Use Case: A support lead asks which of 40 seeded Chinook demo Cases could be handled by an AI agent; the Skill clusters the Cases, quantifies the handling split, and writes a report plus a LangGraph design document. ## Quick Start Analyze the seeded Chinook demo Cases in Salesforce and produce a support demand report with a LangGraph agent design.
