What problem does it solve? New employees who have never used Claude Code, MCP, or AI assistants often feel lost when joining a company that relies on these tools. This Skill provides a complete, structured onboarding program that teaches the company AI workflow architecture from scratch and verifies the learner can work independently before finishing. ## Core Features & Use Cases - Seven-Layer Architecture Teaching: Explains Data/Tools, MCP, Claude Code + Skills, Processing, Output, Human Review, and Evidence using a restaurant analogy, with comprehension check-ins after every layer. - Guided First-Task Practice: Maps the employee's real first task to the architecture, walks them through it, then has them complete a similar task solo. - Formal Assessment & Certification: Runs teach-back and independent-task assessments, issues a readiness decision (READY, RETRY, BLOCKED, etc.), and generates a cheat sheet, learning summary, and completion certificate in onboarding-output/. - Use Case: A non-technical sales hire says "This is my first task and I've never used AI tools." The Skill calibrates to zero knowledge, teaches each layer with check-ins, walks them through generating a sales report, assesses them, and saves their onboarding records. ## Quick Start Say "I'm new here and this is my first task" or invoke /new-joinee with your role to begin the full guided onboarding session.
