# Dental Social Media Ai
## Step-by-Step Training Manual

Version: 2026-05-08
Audience: Founders, office managers, social media coordinators, and ops assistants
Estimated training time: 45-60 minutes

---

## 1. What This System Does
Dental Social Media Ai turns one source asset into daily social output, routes auto-publish payloads, and optionally generates DM drafts for approval/send workflows.

Core outcomes:
- consistent posting cadence
- faster response time to warm leads
- less founder time spent writing day-by-day content

---

## 2. Prerequisites (Before First Login)
Confirm these are ready:

1. App is running locally:
   - `npm start`
   - URL: `http://127.0.0.1:8080`
2. Runtime configured:
   - Claude mode: `CLAUDE_API_KEY` set
   - or webhook runtime configured
3. Integration webhooks configured as needed:
   - `CONTENT_STACK_WEBHOOK_URL`
   - `CONTENT_AUTOPUBLISH_WEBHOOK_URL`
   - `DM_APPROVAL_WEBHOOK_URL`
   - `DM_SEND_WEBHOOK_URL`
4. Social profile links available:
   - LinkedIn, X, Instagram, Facebook, and any optional channels

Pass criteria:
- `/api/health` returns `ok: true`

---

## 3. Product Tour (5 Minutes)
Use these pages:

1. Main dashboard: `index.html`
   - overview cards and pricing tiers
2. Configure + Run: `run.html?agent=content_flywheel`
   - manual run setup
3. Visual Ops dashboard: `visual-dashboard.html`
   - live integrations, autopilot controls, job monitor, report snapshot
4. Report page: `report.html?job=<jobId>&agent=content_flywheel`
   - final output and DM drafts

---

## 4. First Manual Run (Hands-On)
Goal: produce one real content pack and report.

Step-by-step:

1. Open `run.html?agent=content_flywheel`
2. Fill required fields:
   - `Seed Asset Type`
   - `Seed Asset Summary`
   - `Auto Publish` (`yes` or `no`)
   - `DM Mode` (`yes` or `no`)
   - `DM Auto Send` (`yes` or `no`)
   - `Daily Autopilot` (`yes` or `no`)
3. Add channel links (at minimum LinkedIn + one secondary channel)
4. Click `LAUNCH AGENT`
5. Wait for loading flow to complete and redirect to report

Pass criteria:
- report loads successfully
- summary + KPI block populated
- DM drafts appear if DM mode was enabled

---

## 5. Report Review Workflow
Use `report.html` after each run.

Review checklist:

1. Summary quality
   - clear narrative
   - correct offer and CTA
2. KPI sanity
   - output aligns to weekly cadence goals
3. Output blocks
   - platform-specific positioning is coherent
4. DM drafts
   - tone and compliance match brand guidelines
5. Action items
   - assign owner and due date to each item

Approval rules:
- approve if messaging is accurate and publish-ready
- rerun if factual errors, wrong tone, or weak CTA are found

---

## 6. Configure Daily Autopilot
Goal: schedule consistent output without manual daily setup.

Step-by-step:

1. Open `visual-dashboard.html`
2. In `Autopilot Controls`, set:
   - `Autopilot Enabled` = on
   - `Time (Local)` = preferred daily post prep time
   - `Auto Publish` = on/off
   - `DM Mode` and `DM Channels`
   - brand and audience fields
   - `Seed Asset Summary` and `Context`
3. Click `SAVE AUTOPILOT`
4. Confirm `Autopilot State = ENABLED`
5. Optional: click `RUN NOW` for immediate validation

Pass criteria:
- save returns success
- new run creates a valid job ID
- job completes and report snapshot updates

---

## 7. DM Mode Operations
Two supported patterns:

1. Draft + approval (recommended)
   - `DM Mode = yes`
   - `DM Auto Send = no`
   - team approves before sending
2. Auto-send (advanced)
   - `DM Mode = yes`
   - `DM Auto Send = yes`
   - only use when compliance and templates are stable

Daily DM SOP:

1. Review DM drafts in report or approval queue
2. Edit for context as needed
3. Approve/send by channel
4. Log responses and booked conversations

---

## 8. Visual Dashboard Daily Routine (10 Minutes)
Do this every business day:

1. Open `visual-dashboard.html`
2. Check KPI row:
   - runtime mode
   - autopilot state
   - webhook coverage
3. Check integration matrix:
   - resolve anything marked `missing`
4. Review live/recent jobs
5. Confirm latest report snapshot quality
6. Trigger `RUN NOW` if today’s output is missing

---

## 9. Weekly Founder Workflow (30 Minutes)
Run this once per week:

1. Prepare one source asset (memo/transcript/call notes)
2. Update `Seed Asset Summary` in autopilot config
3. Set this week’s CTA and campaign context
4. Run one supervised batch
5. Review report + DM drafts
6. Adjust platform priority and DM tone based on performance

Weekly KPI targets:
- posting consistency (target 5-7 posts/week)
- response speed to warm leads
- qualified conversations started via DM

---

## 10. Troubleshooting

### Problem: Run fails immediately
Checks:
1. confirm server is running
2. check `CLAUDE_API_KEY` or runtime webhook settings
3. test `/api/health`

### Problem: Autopublish not firing
Checks:
1. `Auto Publish` must be `yes`
2. `CONTENT_AUTOPUBLISH_WEBHOOK_URL` configured
3. downstream n8n workflow active

### Problem: DM drafts missing
Checks:
1. `DM Mode` must be `yes`
2. DM channels must be valid (`instagram,facebook,x,linkedin`)
3. runtime report returned `dmDrafts`

### Problem: Dashboard shows missing integrations
Checks:
1. verify env vars in app and n8n
2. restart app after env changes
3. confirm webhook endpoints are reachable

---

## 11. Team Roles and Responsibilities

1. Founder / Lead dentist
   - defines weekly message and offer
   - approves campaign tone
2. Operator / VA
   - updates source summary
   - monitors dashboard and run health
3. Reviewer
   - approves posts/DM drafts
   - enforces compliance standards

---

## 12. Onboarding Script (Use With New Staff)
Use this 5-step script:

1. "Today you’ll learn how one source becomes a week of social content."
2. "First we run manually so you can inspect quality before automation."
3. "Then we enable autopilot and verify it from the visual dashboard."
4. "DM mode stays approval-first until you’re comfortable with quality."
5. "Every day you do a 10-minute dashboard check and resolve red flags fast."

---

## 13. Certification Checklist (Internal)
Mark complete when trainee can:

1. launch a manual run without assistance
2. configure and save autopilot schedule
3. trigger run-now and monitor job status
4. review report KPIs and DM drafts
5. identify and resolve one integration issue

When all five are complete, trainee is production-ready.

