# Video Walkthrough Scripts These are short first-version scripts for a GitHub Pages + PDF workbook launch. Record as screen-share videos. Keep each video 5–8 minutes. --- ## Video 1 — Welcome: A System with the Heart of a Nurse **Goal:** Explain the course promise and nursing posture through the Florence Nightingale frame. Opening: > Florence Nightingale carried a lamp — and a notebook. She stayed present with the suffering, and she charted the data that proved most deaths were preventable. Heart and evidence, together. This course hands you both instruments for the AI age. It is not about replacing nursing judgment with AI. It is about building an operating system that carries the busywork so you can carry the patient. Cover: 1. Florence's two instruments: the caring heart and the counting mind 2. Who this is for — student nurse, staff nurse, nurse leader — and the ache each one carries 3. The three spheres: personal, professional, community / entrepreneurship 4. Core rule: Hermes supports. Humans judge. Nurses steward. 5. What this course is not: no PHI, no patient-specific care decisions, no replacement for licensed judgment 6. What students will build by the end Close: > Your first task is not to install a tool. Your first task is to define the values that will guide the tool. Carry the lamp. Keep the ledger. --- ## Video 2 — Install Hermes Desktop and Set Safe Boundaries **Goal:** Walk through installation and safe orientation. Cover: 1. Download/install Hermes Desktop from the official Hermes Agent site or official Nous Research release path 2. On macOS: open the `.dmg`, drag Hermes into Applications, eject the DMG, and launch from Applications or Spotlight 3. Complete first-run onboarding 4. Open the UI: sessions, pinned messages, artifacts, skills, tools, messaging, settings 5. Pick a provider/model; use Nous Portal or another supported provider if available, and remember that free/low-cost model availability can change 6. Optional Terminal verification: `hermes --tui` 7. Set the workspace to the course folder, not the whole computer 8. Protect API keys: never place keys in public notes, screenshots, GitHub, or course forums 9. Keep manual approval on for beginner use 10. Explain memory as useful but governed: no PHI, no secrets, no confidential employer data 11. Create a nurse-centered workspace folder 12. Add `NO-PHI-BOUNDARY.md` 13. Add `HUMAN-AGENCY-RULES.md` 14. Run first safe prompt 15. Demonstrate one bounded file task inside the course folder 16. Create separated profiles: Personal Projects, Personal Learning, and Professional Non-PHI 17. Explain that profiles separate config, SOUL, memory, sessions, skills, cron jobs, and identity — but do not replace privacy governance Safe first prompt: ```text Read my Nurse-AI-OS folder and summarize the starter files you see. Do not access other folders. Do not modify anything yet. ``` Safe file-task prompt: ```text Create a new folder called Practice inside My-Nurse-AI-OS and write a short README explaining that this is a safe test folder. ``` Safety line: > If it could identify a patient, do not paste it. If it could affect care, do not outsource it. --- ## Video 3 — Build Your Nurse-Centered Chief of Staff **Goal:** Create the personal mission-control workflow. Cover: 1. Daily brief prompt 2. Weekly review prompt 3. Personal dashboard 4. Energy and recovery protection 5. Decision ownership Demo prompt: ```text Act as my nurse-centered Chief of Staff. Help me identify what matters today, what can wait, what needs follow-up, and what action protects my wellbeing. ``` --- ## Video 4 — Build Your Learning OS **Goal:** Show how AI supports learning without replacing expertise. Cover: 1. Choose one topic 2. Build a pathway 3. Create practice scenarios 4. Generate self-test questions 5. Identify what evidence to review 6. Identify what a human expert should verify Key line: > A Learning OS should build your capability, not your dependency. --- ## Video 5 — Build Your Professional + Community OS **Goal:** Map career and community/entrepreneurial support. Cover: 1. Career map 2. Certification plan 3. Research review 4. Idea map 5. Community needs map 6. Smallest safe pilot Entrepreneurship boundary: > Be imaginative, but not reckless. Do not promise what you have not proven. --- ## Video 6 — Govern, Loop, and Sustain **Goal:** Teach HERMES, EDENA, and safe loops. Cover: 1. HERMES Transformation Protocol 2. EDENA Stewardship Lens 3. Loop Charter 4. Stop conditions 5. Monthly AI OS audit 6. Integration Gate for Obsidian, Google Workspace, GitHub, and GitHub Pages 7. macOS GitHub setup: `brew install gh`, `gh auth login`, private repo first, public Pages only when publishable 8. Image generation for learning dashboards, personal sites, and public-safe visual drafts 9. Knowledge graph dashboard prompt pack: static image, ambient loop, and image-to-video patterns 10. Video generation for short 4–8 second ambient dashboard loops 11. Three-lane research setup: personal knowledge, academic/medical reading, and technical project research 12. Perplexity MCP as an optional citation-rich source, never as final truth 13. Messaging gateway comparison: Telegram, WhatsApp, Slack, Discord, Signal, SMS, and Email 14. Team Lab profile blueprint: SOUL.md, allowlisted gateway, shared skills, cron reports, and launch checklist 15. Screenshot workflows: paste/attach images, browser_vision for public/test pages, and no-PHI visual boundaries 16. macOS Computer Use + Magnet layouts for study, writing, coding, and meeting setups 17. AI browser boundary: Hermes as hub, Comet/Atlas as optional manual research browsers unless governed 18. YouTube transcripts into Hermes: lecture → timestamped chapters, Obsidian note, follow-up questions 19. Living and thriving with AI: cognitive load, ergonomics, relationships, critical thinking, systems thinking, and leader/educator practice 20. 90-minute workshop outline for nurses, leaders, educators, or students 21. Google Workspace as a governed suite: Gmail, Drive, Docs, Sheets, Slides, Meet, Calendar, and core services under BAA only when eligible/configured/approved 22. Safe AI with PHI: personal Gmail and generic AI are no-PHI; use fictional/de-identified/public/synthetic material unless explicitly approved 23. Why every connector, media tool, messaging channel, screenshot, browser, Workspace service, and MCP server needs least privilege, cost awareness, approval gates, and a human owner Key line: > Every loop needs a leash, a log, a limit, and a human owner. Close: > The goal is not a smarter machine. The goal is a stronger nurse.