What problem does it solve? AI agents without intentional persona design sound generic, inconsistent, or off-brand, which erodes user trust and adoption. This Skill turns brand guides, URLs, or plain descriptions into a structured, reviewable persona document and encodes it into Agentforce configuration. ## Core Features & Use Cases - Persona Design Workflow: Drafts identity traits, 12 dimensions across 5 categories (Register, Voice, Tone, Delivery, Chatting Style), phrase books, never-say lists, tone flex rules, and lexicons from any starting input, then validates with sample dialog. - 50-Point Scoring Rubric: Evaluates persona quality across identity coherence, dimension consistency, behavioral specificity, phrase book quality, and sample quality on request. - Agentforce Encoding: Converts a finished persona into Agent Script YAML blocks (system.instructions, per-topic reasoning.instructions, loading text) or Agentforce Builder field values (Name, Role, Company, Welcome, Error), with telephony adjustments when needed. - Use Case: A team launching a customer-facing support agent provides their brand guide PDF; the Skill drafts a complete persona, shows sample dialog for refinement, scores it, and outputs ready-to-paste Agent Script YAML. ## Quick Start Invoke the sf-ai-agentforce-persona skill and provide a brand guide, URL, or text description of the agent you want to design.