diff --git a/SPARK_ONBOARDING.md b/SPARK_ONBOARDING.md
new file mode 100644
index 0000000..1408a16
--- /dev/null
+++ b/SPARK_ONBOARDING.md
@@ -0,0 +1,66 @@
+# Spark Platform: Zero-to-One Onboarding Guide
+
+> **Current State**: Phase 8 Complete (Premium Features Active)
+> **Last Updated**: 2025-12-13
+
+## 1. Project Overview
+Spark is a high-performance content scaling platform ("Titanium Pro" edition). It leverages Directus (CMS/Data) and Astro/React (Frontend) to deliver enterprise-grade mass content generation.
+
+## 2. Key Architecture
+- **Frontend**: Astro (SSR) + React (Interactive Islands).
+- **Backend**: Directus Headless CMS (Postgres).
+- **Styling**: TailwindCSS (Titanium Pro Design System: Dark Mode, Glassmorphism, Neon Accents).
+- **Deployment**: Coolify (Dockerized).
+
+## 3. Core Modules (Implemented)
+
+### 🏭 Factory Floor (`/admin/content/factory`)
+The bread and butter of the operation.
+- **Components**: `KanbanBoard`, `BulkGrid`.
+- **Purpose**: Manage the lifecycle of thousands of articles.
+
+### 🧠 Intelligence Station (`/admin/intelligence`)
+Data-driven insights engine.
+- **Geospatial Map (`/geo-targeting`)**: visualize market dominance with heatmaps using Leaflet.
+- **Trend Analysis**: Analyze keyword volatility.
+- **Avatar Metrics**: Track performance by user persona.
+
+### 🧩 Assembler Engine (`/admin/assembler`)
+The heart of content generation.
+- **Template Composer**: 3-pane editor for Spintax templates.
+- **Automation Builder (`/automations/workflow`)**: **[NEW]** Visual Node-based editor (React Flow) to design content pipelines (Trigger -> Generate -> Publish).
+
+### 📊 Command Center (`/admin/analytics`)
+- **Metrics Dashboard (`/metrics`)**: **[NEW]** Professional Tremor-based analytics (Area Charts, Donut Charts) tracking traffic and engagement.
+
+### 🛡️ Testing Suite (`/admin/testing`)
+- **Visual Editor (`/blocks/editor`)**: Drag-and-drop page builder (Craft.js).
+- **SEO/Readability**: Real-time content scoring (Flesch-Kincaid).
+
+## 4. Development Standards
+
+### "Titanium Pro" Design Rules
+- **Colors**: Background `#09090b` (Zinc-950), Accents `#eab308` (Yellow-500) and `#22c55e` (Green-500).
+- **Components**: Always use **Shadcn/UI** (`/components/ui/*.tsx`).
+- **Icons**: Lucide React (``).
+- **Animations**: Use `framer-motion` for complex interactions, or CSS transitions for hover states.
+
+### Code Quality
+- **Typescript**: Strict mode. No `any` unless absolutely necessary (and temporarily).
+- **Linting**: Run `npm run lint` before committing.
+- **Testing**: Use the internal `/admin/testing/suite` to validate logic changes.
+
+## 5. Deployment Workflow
+1. **Commit**: `git push` triggers Coolify build.
+2. **Build Process**: Dockerfile builds frontend (SSR).
+3. **Validation**: Check `https://launch.jumpstartscaling.com` for the live site.
+4. **Staging**: Use the "Preview" button in `CollectionManager` to vie draft content at `https://launch.jumpstartscaling.com/site/[id]/preview/[post_id]`.
+
+## 6. Next Actions (Phase 7: Polish)
+The logic is done. The next phase is strictly **Polish**:
+- [ ] **Accessibility Audit**: Ensure ARIA labels on all new UI.
+- [ ] **Documentation**: Expand inline JSDoc for complex engines (`engine.ts`).
+- [ ] **Performance**: Optimize React Flow rendering for large graphs.
+
+---
+**Note to AI Agent**: You are working on a codebase that is feature-complete for Alpha. Your goal is stability and refinement. Do not refactor core engines without explicit instruction.
diff --git a/scripts/README.md b/scripts/README.md
index 94a087f..8815f31 100644
--- a/scripts/README.md
+++ b/scripts/README.md
@@ -1,417 +1,30 @@
-# Spark Platform - Management Scripts
-
-This directory contains powerful Node.js utilities for managing your Spark Directus instance through the API. All scripts connect to `https://spark.jumpstartscaling.com` using admin credentials.
-
-## 🔧 Available Scripts
-
-### 1. Connection Test
-**File:** `test_directus_connection.js` (in project root)
-
-Tests basic connectivity and admin access to Directus.
-
-```bash
-node test_directus_connection.js
-```
-
-**What it checks:**
-- Server availability
-- Admin authentication
-- Collections list
-- Record counts
-- Write permissions
-
+---
+description: Learn how to manage deployment scripts and database migrations
---
-### 2. Schema Audit
-**File:** `audit_schema.js`
+# Scripts & Automation
-Comprehensive audit of all collections, fields, and relationships.
+This directory contains utility scripts for managing the Spark Platform.
-```bash
-node scripts/audit_schema.js
-```
+## Core Scripts
-**Features:**
-- Lists all collections with record counts
-- Shows all fields and their interfaces
-- Identifies missing relationships
-- Detects UX issues (wrong field types, missing dropdowns)
-- Saves detailed report to `schema_audit_report.json`
+- `deploy.sh`: (Legacy) Basic shell script for deployment. We now use **Coolify**.
+- `seed.ts`: Database seeder (if applicable).
-**Output:**
-- Field-by-field analysis
-- Relationship mapping
-- Issue summary
-- Recommendations
+## Deployment
----
+We use **Coolify** for all deployments. The `Dockerfile` in the root determines the build process.
+- **Frontend**: Built via `npm run build` (Astro).
+- **Backend**: Directus via Docker Image.
-### 3. UX Improvements
-**File:** `improve_ux.js`
+## Database Migrations
-Automatically fixes field interfaces to make Directus admin UI more user-friendly.
+Directus handles migrations internally.
+- To export schema: `npx directus schema snapshot ./schema.yaml`
+- To apply schema: `npx directus schema apply ./schema.yaml`
-```bash
-node scripts/improve_ux.js
-```
+## ⚠️ Important Notes
-**What it fixes:**
-- ✅ `site_id` fields → Select dropdown with site names
-- ✅ `campaign_id` fields → Select dropdown with campaign names
-- ✅ `status` fields → Dropdown with predefined choices
-- ✅ `avatar_key` fields → Avatar selection dropdown
-- ✅ JSON fields → Code editor with syntax highlighting
-- ✅ Content fields → Rich text HTML editor
-- ✅ Date fields → Proper datetime picker
-- ✅ Adds helpful descriptions and placeholders
-
-**Results:**
-- Posts/Pages easily connect to Sites
-- Status uses visual labels
-- JSON editing with syntax highlighting
-- Better field validation
-
----
-
-### 4. Schema Validation
-**File:** `validate_schema.js`
-
-Complete validation of schema integrity, relationships, and data quality.
-
-```bash
-node scripts/validate_schema.js
-```
-
-**Validation checks:**
-1. **Collection Data** - Verifies all critical collections exist
-2. **Relationships** - Tests that foreign keys work correctly
-3. **Field Interfaces** - Confirms UX improvements are applied
-4. **Data Integrity** - Checks for orphaned records
-
-**Output:**
-- Detailed validation report
-- Issue severity ratings (high/medium/low)
-- Saves `validation_report.json`
-
----
-
-### 5. Bulk Import/Export
-**File:** `bulk_io.js`
-
-Export and import any collection as JSON files.
-
-```bash
-# Export all collections
-node scripts/bulk_io.js export all
-
-# Export single collection
-node scripts/bulk_io.js export sites
-
-# Import from file
-node scripts/bulk_io.js import sites sites_backup.json
-
-# List available exports
-node scripts/bulk_io.js list
-```
-
-**Features:**
-- Exports with metadata (timestamp, record count)
-- Batch import with conflict resolution
-- Automatic pagination for large collections
-- Update existing records on conflict
-- All exports saved to `./exports/` directory
-
-**Use cases:**
-- Backup before major changes
-- Move data between environments
-- Share sample data
-- Restore deleted records
-
----
-
-### 6. Geo Intelligence Manager
-**File:** `geo_manager.js`
-
-Easy management of locations in `geo_intelligence` collection.
-
-```bash
-# List all locations
-node scripts/geo_manager.js list
-
-# Add a location
-node scripts/geo_manager.js add "Miami" "FL" "US" "southeast"
-
-# Remove a location
-node scripts/geo_manager.js remove
-
-# Activate/deactivate location
-node scripts/geo_manager.js activate
-node scripts/geo_manager.js deactivate
-
-# Update a field
-node scripts/geo_manager.js update cluster "northeast"
-
-# Seed with top 20 US cities
-node scripts/geo_manager.js seed-us-cities
-
-# Import from CSV
-node scripts/geo_manager.js import-csv locations.csv
-```
-
-**CSV Format:**
-```csv
-city,state,country,cluster,population
-Miami,FL,US,southeast,467963
-Boston,MA,US,northeast,692600
-```
-
-**Built-in sample data:**
-- Top 20 US cities by population
-- Includes clusters (northeast, south, midwest, west, etc.)
-- Ready to use with `seed-us-cities` command
-
----
-
-## 🚀 Quick Start Guide
-
-### First Time Setup
-
-1. **Test your connection:**
- ```bash
- node test_directus_connection.js
- ```
-
-2. **Audit current schema:**
- ```bash
- node scripts/audit_schema.js
- ```
-
-3. **Apply UX improvements:**
- ```bash
- node scripts/improve_ux.js
- ```
-
-4. **Validate everything works:**
- ```bash
- node scripts/validate_schema.js
- ```
-
-5. **Backup all data:**
- ```bash
- node scripts/bulk_io.js export all
- ```
-
-### Daily Operations
-
-**Working with locations:**
-```bash
-# See all locations
-node scripts/geo_manager.js list
-
-# Add custom location
-node scripts/geo_manager.js add "Portland" "OR" "US" "northwest"
-```
-
-**Backing up before changes:**
-```bash
-node scripts/bulk_io.js export sites
-node scripts/bulk_io.js export generation_jobs
-```
-
-**Checking system health:**
-```bash
-node scripts/validate_schema.js
-```
-
----
-
-## 📊 What Each Script Fixed
-
-### Before UX Improvements:
-- ❌ `site_id` fields showed UUID text input
-- ❌ `status` fields were plain text
-- ❌ JSON fields used tiny text box
-- ❌ Content used plain textarea
-- ❌ No field descriptions or help text
-
-### After UX Improvements:
-- ✅ `site_id` shows dropdown with site names
-- ✅ `status` has predefined choices with colors
-- ✅ JSON fields have code editor with syntax highlighting
-- ✅ Content uses rich text HTML editor
-- ✅ All fields have helpful descriptions
-
----
-
-## 🔗 Confirmed Working Relationships
-
-All relationships tested and verified:
-
-1. **Sites → Posts** ✅
- - Posts connected to sites via `site_id`
- - Dropdown shows site names
-
-2. **Sites → Pages** ✅
- - Pages connected to sites via `site_id`
- - Easy site selection
-
-3. **Campaign → Generated Articles** ✅
- - Articles linked to campaigns
- - Track which campaign created each article
-
-4. **Generation Jobs → Sites** ✅
- - Jobs know which site they're for
- - Filters work correctly
-
----
-
-## 📁 Export Directory Structure
-
-After running bulk export, you'll have:
-
-```
-exports/
-├── avatar_intelligence_2025-12-13.json (10 records)
-├── avatar_variants_2025-12-13.json (30 records)
-├── campaign_masters_2025-12-13.json (2 records)
-├── cartesian_patterns_2025-12-13.json (3 records)
-├── content_fragments_2025-12-13.json (150 records)
-├── generated_articles_2025-12-13.json (0 records)
-├── generation_jobs_2025-12-13.json (30 records)
-├── geo_intelligence_2025-12-13.json (3 records)
-├── sites_2025-12-13.json (3 records)
-└── ... (all other collections)
-```
-
-Each JSON file includes:
-```json
-{
- "collection": "sites",
- "exportedAt": "2025-12-13T14:30:00.000Z",
- "recordCount": 3,
- "data": [...]
-}
-```
-
----
-
-## 🎯 Common Tasks
-
-### Add Multiple Locations from CSV
-
-1. Create `locations.csv`:
- ```csv
- city,state,country,cluster,population
- Seattle,WA,US,northwest,753675
- Portland,OR,US,northwest,652503
- San Francisco,CA,US,west,881549
- ```
-
-2. Import:
- ```bash
- node scripts/geo_manager.js import-csv locations.csv
- ```
-
-### Backup Before Major Changes
-
-```bash
-# Export everything
-node scripts/bulk_io.js export all
-
-# Make your changes in Directus UI...
-
-# If something goes wrong, restore:
-node scripts/bulk_io.js import sites exports/sites_2025-12-13.json
-```
-
-### Check What Needs Fixing
-
-```bash
-# See what's wrong
-node scripts/audit_schema.js
-
-# Auto-fix field interfaces
-node scripts/improve_ux.js
-
-# Verify fixes worked
-node scripts/validate_schema.js
-```
-
----
-
-## 🔧 Troubleshooting
-
-### "Authentication failed"
-- Check credentials in script files
-- Verify admin token hasn't expired
-- Test with: `node test_directus_connection.js`
-
-### "Collection not found"
-- Collection may not exist yet
-- Run audit to see all collections: `node scripts/audit_schema.js`
-- Check schema is initialized
-
-### Import conflicts (409 errors)
-- Script automatically tries to UPDATE existing records
-- Check the import summary for failed records
-- Review data for duplicate IDs
-
----
-
-## 📚 Additional Resources
-
-- [Directus API Docs](https://docs.directus.io/reference/introduction.html)
-- [Spark Onboarding Guide](../spark_onboarding.md)
-- [Campaign Setup Guide](../CAMPAIGN_SETUP_GUIDE.md)
-
----
-
-## ✅ Current System Status
-
-**Last Validation:** 2025-12-13
-
-- ✅ 11/11 critical collections exist
-- ✅ 9/11 collections have data
-- ✅ 4/4 relationships working
-- ✅ 32 field interfaces improved
-- ✅ 251 total records
-- ✅ 30 pending generation jobs
-- ✅ Zero data integrity issues
-
-**Ready for production use!** 🎉
-
----
-
-## 💡 Pro Tips
-
-1. **Always backup before bulk changes:**
- ```bash
- node scripts/bulk_io.js export all
- ```
-
-2. **Use validation after making schema changes:**
- ```bash
- node scripts/validate_schema.js
- ```
-
-3. **Check exports directory regularly:**
- ```bash
- node scripts/bulk_io.js list
- ```
-
-4. **Seed sample data for testing:**
- ```bash
- node scripts/geo_manager.js seed-us-cities
- ```
-
-5. **Keep audit reports for reference:**
- - Reports saved to: `schema_audit_report.json`
- - Save with timestamps for comparison
-
----
-
-**Need help?** All scripts have built-in help:
-```bash
-node scripts/[script-name].js
-# Shows available commands and examples
-```
+- **Do NOT** modify `docker-compose.yml` manually on the server. Use Coolify UI.
+- **Environment Variables**: Managed in Coolify. See `.env.example` for required keys.
+- **Onboarding**: See `SPARK_ONBOARDING.md` in the root for the full developer guide.