jeffemmett
/
canvas-website
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Merge origin/main into feature/google-export 

Bring in all the latest changes from main including:
- Index validation and migration for tldraw shapes
- UserSettingsModal with integrations tab
- CryptID authentication updates
- AI services (image gen, video gen, mycelial intelligence)
- Automerge sync improvements
- Various UI improvements

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
This commit is contained in:
Jeff Emmett 2025-12-04 16:29:34 -08:00
parent 58ff544c46 7ef0533a8f
commit 8411211ca6
195 changed files with 35812 additions and 10230 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
.cfignore Normal file
View File

@ -0,0 +1,4 @@
# Ignore Cloudflare Worker configuration files during Pages deployment
# These are only used for separate Worker deployments
worker/
*.toml

14
.env.example
View File

@ -4,10 +4,22 @@ VITE_GOOGLE_MAPS_API_KEY='your_google_maps_api_key'
VITE_DAILY_DOMAIN='your_daily_domain' VITE_DAILY_DOMAIN='your_daily_domain'
VITE_TLDRAW_WORKER_URL='your_worker_url' VITE_TLDRAW_WORKER_URL='your_worker_url'
# AI Configuration
# AI Orchestrator with Ollama (FREE local AI - highest priority)
VITE_OLLAMA_URL='https://ai.jeffemmett.com'
# RunPod API (Primary AI provider when Ollama unavailable)
# Users don't need their own API keys - RunPod is pre-configured
VITE_RUNPOD_API_KEY='your_runpod_api_key_here'
VITE_RUNPOD_TEXT_ENDPOINT_ID='your_text_endpoint_id' # vLLM for chat/text
VITE_RUNPOD_IMAGE_ENDPOINT_ID='your_image_endpoint_id' # Automatic1111/SD
VITE_RUNPOD_VIDEO_ENDPOINT_ID='your_video_endpoint_id' # Wan2.2
VITE_RUNPOD_WHISPER_ENDPOINT_ID='your_whisper_endpoint_id' # WhisperX
# Worker-only Variables (Do not prefix with VITE_) # Worker-only Variables (Do not prefix with VITE_)
CLOUDFLARE_API_TOKEN='your_cloudflare_token' CLOUDFLARE_API_TOKEN='your_cloudflare_token'
CLOUDFLARE_ACCOUNT_ID='your_account_id' CLOUDFLARE_ACCOUNT_ID='your_account_id'
CLOUDFLARE_ZONE_ID='your_zone_id' CLOUDFLARE_ZONE_ID='your_zone_id'
R2_BUCKET_NAME='your_bucket_name' R2_BUCKET_NAME='your_bucket_name'
R2_PREVIEW_BUCKET_NAME='your_preview_bucket_name' R2_PREVIEW_BUCKET_NAME='your_preview_bucket_name'
DAILY_API_KEY=your_daily_api_key_here DAILY_API_KEY=your_daily_api_key_here

28
.github/workflows/mirror-to-gitea.yml vendored Normal file
View File

@ -0,0 +1,28 @@
name: Mirror to Gitea
on:
push:
branches:
- main
- master
workflow_dispatch:
jobs:
mirror:
runs-on: ubuntu-latest
steps:
- name: Checkout repository
uses: actions/checkout@v4
with:
fetch-depth: 0
- name: Mirror to Gitea
env:
GITEA_TOKEN: ${{ secrets.GITEA_TOKEN }}
GITEA_USERNAME: ${{ secrets.GITEA_USERNAME }}
run: |
REPO_NAME=$(basename $GITHUB_REPOSITORY)
git remote add gitea https://$GITEA_USERNAME:$GITEA_TOKEN@gitea.jeffemmett.com/jeffemmett/$REPO_NAME.git || true
git push gitea --all --force
git push gitea --tags --force

1
.gitignore vendored
View File

@ -175,3 +175,4 @@ dist
.env.*.local .env.*.local
.dev.vars .dev.vars
.env.production .env.production
.aider*

2
.npmrc
View File

@ -1,3 +1,3 @@
legacy-peer-deps=true legacy-peer-deps=true
strict-peer-dependencies=false strict-peer-dependencies=false
auto-install-peers=true auto-install-peers=true

1
.nvmrc Normal file
View File

@ -0,0 +1 @@
20

626
AI_SERVICES_DEPLOYMENT_GUIDE.md Normal file
View File

@ -0,0 +1,626 @@
# AI Services Deployment & Testing Guide
Complete guide for deploying and testing the AI services integration in canvas-website with Netcup RS 8000 and RunPod.
---
## 🎯 Overview
This project integrates multiple AI services with smart routing:
**Smart Routing Strategy:**
- **Text/Code (70-80% workload)**: Local Ollama on RS 8000 → **FREE**
- **Images - Low Priority**: Local Stable Diffusion on RS 8000 → **FREE** (slow ~60s)
- **Images - High Priority**: RunPod GPU (SDXL) → **$0.02/image** (fast ~5s)
- **Video Generation**: RunPod GPU (Wan2.1) → **$0.50/video** (30-90s)
**Expected Cost Savings:** $86-350/month compared to persistent GPU instances
---
## 📦 What's Included
### AI Services:
1. ✅ **Text Generation (LLM)**
- RunPod integration via `src/lib/runpodApi.ts`
- Enhanced LLM utilities in `src/utils/llmUtils.ts`
- AI Orchestrator client in `src/lib/aiOrchestrator.ts`
- Prompt shapes, arrow LLM actions, command palette
2. ✅ **Image Generation**
- ImageGenShapeUtil in `src/shapes/ImageGenShapeUtil.tsx`
- ImageGenTool in `src/tools/ImageGenTool.ts`
- Mock mode **DISABLED** (ready for production)
- Smart routing: low priority → local CPU, high priority → RunPod GPU
3. ✅ **Video Generation (NEW!)**
- VideoGenShapeUtil in `src/shapes/VideoGenShapeUtil.tsx`
- VideoGenTool in `src/tools/VideoGenTool.ts`
- Wan2.1 I2V 14B 720p model on RunPod
- Always uses GPU (no local option)
4. ✅ **Voice Transcription**
- WhisperX integration via `src/hooks/useWhisperTranscriptionSimple.ts`
- Automatic fallback to local Whisper model
---
## 🚀 Deployment Steps
### Step 1: Deploy AI Orchestrator on Netcup RS 8000
**Prerequisites:**
- SSH access to Netcup RS 8000: `ssh netcup`
- Docker and Docker Compose installed
- RunPod API key
**1.1 Create AI Orchestrator Directory:**
```bash
ssh netcup << 'EOF'
mkdir -p /opt/ai-orchestrator/{services/{router,workers,monitor},configs,data/{redis,postgres,prometheus}}
cd /opt/ai-orchestrator
EOF
```
**1.2 Copy Configuration Files:**
From your local machine, copy the AI orchestrator files created in `NETCUP_MIGRATION_PLAN.md`:
```bash
# Copy docker-compose.yml
scp /path/to/docker-compose.yml netcup:/opt/ai-orchestrator/
# Copy service files
scp -r /path/to/services/* netcup:/opt/ai-orchestrator/services/
```
**1.3 Configure Environment Variables:**
```bash
ssh netcup "cat > /opt/ai-orchestrator/.env" << 'EOF'
# PostgreSQL
POSTGRES_PASSWORD=$(openssl rand -hex 16)
# RunPod API Keys
RUNPOD_API_KEY=your_runpod_api_key_here
RUNPOD_TEXT_ENDPOINT_ID=your_text_endpoint_id
RUNPOD_IMAGE_ENDPOINT_ID=your_image_endpoint_id
RUNPOD_VIDEO_ENDPOINT_ID=your_video_endpoint_id
# Grafana
GRAFANA_PASSWORD=$(openssl rand -hex 16)
# Monitoring
ALERT_EMAIL=your@email.com
COST_ALERT_THRESHOLD=100
EOF
```
**1.4 Deploy the Stack:**
```bash
ssh netcup << 'EOF'
cd /opt/ai-orchestrator
# Start all services
docker-compose up -d
# Check status
docker-compose ps
# View logs
docker-compose logs -f router
EOF
```
**1.5 Verify Deployment:**
```bash
# Check health endpoint
ssh netcup "curl http://localhost:8000/health"
# Check API documentation
ssh netcup "curl http://localhost:8000/docs"
# Check queue status
ssh netcup "curl http://localhost:8000/queue/status"
```
### Step 2: Setup Local AI Models on RS 8000
**2.1 Download Ollama Models:**
```bash
ssh netcup << 'EOF'
# Download recommended models
docker exec ai-ollama ollama pull llama3:70b
docker exec ai-ollama ollama pull codellama:34b
docker exec ai-ollama ollama pull deepseek-coder:33b
docker exec ai-ollama ollama pull mistral:7b
# Verify
docker exec ai-ollama ollama list
# Test a model
docker exec ai-ollama ollama run llama3:70b "Hello, how are you?"
EOF
```
**2.2 Download Stable Diffusion Models:**
```bash
ssh netcup << 'EOF'
mkdir -p /data/models/stable-diffusion/sd-v2.1
cd /data/models/stable-diffusion/sd-v2.1
# Download SD 2.1 weights
wget https://huggingface.co/stabilityai/stable-diffusion-2-1/resolve/main/v2-1_768-ema-pruned.safetensors
# Verify
ls -lh v2-1_768-ema-pruned.safetensors
EOF
```
**2.3 Download Wan2.1 Video Generation Model:**
```bash
ssh netcup << 'EOF'
# Install huggingface-cli
pip install huggingface-hub
# Download Wan2.1 I2V 14B 720p
mkdir -p /data/models/video-generation
cd /data/models/video-generation
huggingface-cli download Wan-AI/Wan2.1-I2V-14B-720P \
--include "*.safetensors" \
--local-dir wan2.1_i2v_14b
# Check size (~28GB)
du -sh wan2.1_i2v_14b
EOF
```
**Note:** The Wan2.1 model will be deployed to RunPod, not run locally on CPU.
### Step 3: Setup RunPod Endpoints
**3.1 Create RunPod Serverless Endpoints:**
Go to [RunPod Serverless](https://www.runpod.io/console/serverless) and create endpoints for:
1. **Text Generation Endpoint** (optional, fallback)
- Model: Any LLM (Llama, Mistral, etc.)
- GPU: Optional (we use local CPU primarily)
2. **Image Generation Endpoint**
- Model: SDXL or SD3
- GPU: A4000/A5000 (good price/performance)
- Expected cost: ~$0.02/image
3. **Video Generation Endpoint**
- Model: Wan2.1-I2V-14B-720P
- GPU: A100 or H100 (required for video)
- Expected cost: ~$0.50/video
**3.2 Get Endpoint IDs:**
For each endpoint, copy the endpoint ID from the URL or endpoint details.
Example: If URL is `https://api.runpod.ai/v2/jqd16o7stu29vq/run`, then `jqd16o7stu29vq` is your endpoint ID.
**3.3 Update Environment Variables:**
Update `/opt/ai-orchestrator/.env` with your endpoint IDs:
```bash
ssh netcup "nano /opt/ai-orchestrator/.env"
# Add your endpoint IDs:
RUNPOD_TEXT_ENDPOINT_ID=your_text_endpoint_id
RUNPOD_IMAGE_ENDPOINT_ID=your_image_endpoint_id
RUNPOD_VIDEO_ENDPOINT_ID=your_video_endpoint_id
# Restart services
cd /opt/ai-orchestrator && docker-compose restart
```
### Step 4: Configure canvas-website
**4.1 Create .env.local:**
In your canvas-website directory:
```bash
cd /home/jeffe/Github/canvas-website-branch-worktrees/add-runpod-AI-API
cat > .env.local << 'EOF'
# AI Orchestrator (Primary - Netcup RS 8000)
VITE_AI_ORCHESTRATOR_URL=http://159.195.32.209:8000
# Or use domain when DNS is configured:
# VITE_AI_ORCHESTRATOR_URL=https://ai-api.jeffemmett.com
# RunPod API (Fallback/Direct Access)
VITE_RUNPOD_API_KEY=your_runpod_api_key_here
VITE_RUNPOD_TEXT_ENDPOINT_ID=your_text_endpoint_id
VITE_RUNPOD_IMAGE_ENDPOINT_ID=your_image_endpoint_id
VITE_RUNPOD_VIDEO_ENDPOINT_ID=your_video_endpoint_id
# Other existing vars...
VITE_GOOGLE_CLIENT_ID=your_google_client_id
VITE_GOOGLE_MAPS_API_KEY=your_google_maps_api_key
VITE_DAILY_DOMAIN=your_daily_domain
VITE_TLDRAW_WORKER_URL=your_worker_url
EOF
```
**4.2 Install Dependencies:**
```bash
npm install
```
**4.3 Build and Start:**
```bash
# Development
npm run dev
# Production build
npm run build
npm run start
```
### Step 5: Register Video Generation Tool
You need to register the VideoGen shape and tool with tldraw. Find where shapes and tools are registered (likely in `src/routes/Board.tsx` or similar):
**Add to shape utilities array:**
```typescript
import { VideoGenShapeUtil } from '@/shapes/VideoGenShapeUtil'
const shapeUtils = [
// ... existing shapes
VideoGenShapeUtil,
]
```
**Add to tools array:**
```typescript
import { VideoGenTool } from '@/tools/VideoGenTool'
const tools = [
// ... existing tools
VideoGenTool,
]
```
---
## 🧪 Testing
### Test 1: Verify AI Orchestrator
```bash
# Test health endpoint
curl http://159.195.32.209:8000/health
# Expected response:
# {"status":"healthy","timestamp":"2025-11-25T12:00:00.000Z"}
# Test text generation
curl -X POST http://159.195.32.209:8000/generate/text \
-H "Content-Type: application/json" \
-d '{
"prompt": "Write a hello world program in Python",
"priority": "normal"
}'
# Expected response:
# {"job_id":"abc123","status":"queued","message":"Job queued on local provider"}
# Check job status
curl http://159.195.32.209:8000/job/abc123
# Check queue status
curl http://159.195.32.209:8000/queue/status
# Check costs
curl http://159.195.32.209:8000/costs/summary
```
### Test 2: Test Text Generation in Canvas
1. Open canvas-website in browser
2. Open browser console (F12)
3. Look for log messages:
- `✅ AI Orchestrator is available at http://159.195.32.209:8000`
4. Create a Prompt shape or use arrow LLM action
5. Enter a prompt and submit
6. Verify response appears
7. Check console for routing info:
- Should see `Using local Ollama (FREE)`
### Test 3: Test Image Generation
**Low Priority (Local CPU - FREE):**
1. Use ImageGen tool from toolbar
2. Click on canvas to create ImageGen shape
3. Enter prompt: "A beautiful mountain landscape"
4. Select priority: "Low"
5. Click "Generate"
6. Wait 30-60 seconds
7. Verify image appears
8. Check console: Should show `Using local Stable Diffusion CPU`
**High Priority (RunPod GPU - $0.02):**
1. Create new ImageGen shape
2. Enter prompt: "A futuristic city at sunset"
3. Select priority: "High"
4. Click "Generate"
5. Wait 5-10 seconds
6. Verify image appears
7. Check console: Should show `Using RunPod SDXL`
8. Check cost: Should show `~$0.02`
### Test 4: Test Video Generation
1. Use VideoGen tool from toolbar
2. Click on canvas to create VideoGen shape
3. Enter prompt: "A cat walking through a garden"
4. Set duration: 3 seconds
5. Click "Generate"
6. Wait 30-90 seconds
7. Verify video appears and plays
8. Check console: Should show `Using RunPod Wan2.1`
9. Check cost: Should show `~$0.50`
10. Test download button
### Test 5: Test Voice Transcription
1. Use Transcription tool from toolbar
2. Click to create Transcription shape
3. Click "Start Recording"
4. Speak into microphone
5. Click "Stop Recording"
6. Verify transcription appears
7. Check if using RunPod or local Whisper
### Test 6: Monitor Costs and Performance
**Access monitoring dashboards:**
```bash
# API Documentation
http://159.195.32.209:8000/docs
# Queue Status
http://159.195.32.209:8000/queue/status
# Cost Tracking
http://159.195.32.209:3000/api/costs/summary
# Grafana Dashboard
http://159.195.32.209:3001
# Default login: admin / admin (change this!)
```
**Check daily costs:**
```bash
curl http://159.195.32.209:3000/api/costs/summary
```
Expected response:
```json
{
"today": {
"local": 0.00,
"runpod": 2.45,
"total": 2.45
},
"this_month": {
"local": 0.00,
"runpod": 45.20,
"total": 45.20
},
"breakdown": {
"text": 0.00,
"image": 12.50,
"video": 32.70,
"code": 0.00
}
}
```
---
## 🐛 Troubleshooting
### Issue: AI Orchestrator not available
**Symptoms:**
- Console shows: `⚠️ AI Orchestrator configured but not responding`
- Health check fails
**Solutions:**
```bash
# 1. Check if services are running
ssh netcup "cd /opt/ai-orchestrator && docker-compose ps"
# 2. Check logs
ssh netcup "cd /opt/ai-orchestrator && docker-compose logs -f router"
# 3. Restart services
ssh netcup "cd /opt/ai-orchestrator && docker-compose restart"
# 4. Check firewall
ssh netcup "sudo ufw status"
ssh netcup "sudo ufw allow 8000/tcp"
```
### Issue: Image generation fails with "No output found"
**Symptoms:**
- Job completes but no image URL returned
- Error: `Job completed but no output data found`
**Solutions:**
1. Check RunPod endpoint configuration
2. Verify endpoint handler returns correct format:
```json
{"output": {"image": "base64_or_url"}}
```
3. Check endpoint logs in RunPod console
4. Test endpoint directly with curl
### Issue: Video generation timeout
**Symptoms:**
- Job stuck in "processing" state
- Timeout after 120 attempts
**Solutions:**
1. Video generation takes 30-90 seconds, ensure patience
2. Check RunPod GPU availability (might be cold start)
3. Increase timeout in VideoGenShapeUtil if needed
4. Check RunPod endpoint logs for errors
### Issue: High costs
**Symptoms:**
- Monthly costs exceed budget
- Too many RunPod requests
**Solutions:**
```bash
# 1. Check cost breakdown
curl http://159.195.32.209:3000/api/costs/summary
# 2. Review routing decisions
curl http://159.195.32.209:8000/queue/status
# 3. Adjust routing thresholds
# Edit router configuration to prefer local more
ssh netcup "nano /opt/ai-orchestrator/services/router/main.py"
# 4. Set cost alerts
ssh netcup "nano /opt/ai-orchestrator/.env"
# COST_ALERT_THRESHOLD=50 # Alert if daily cost > $50
```
### Issue: Local models slow or failing
**Symptoms:**
- Text generation slow (>30s)
- Image generation very slow (>2min)
- Out of memory errors
**Solutions:**
```bash
# 1. Check system resources
ssh netcup "htop"
ssh netcup "free -h"
# 2. Reduce model size
ssh netcup << 'EOF'
# Use smaller models
docker exec ai-ollama ollama pull llama3:8b # Instead of 70b
docker exec ai-ollama ollama pull mistral:7b # Lighter model
EOF
# 3. Limit concurrent workers
ssh netcup "nano /opt/ai-orchestrator/docker-compose.yml"
# Reduce worker replicas if needed
# 4. Increase swap (if low RAM)
ssh netcup "sudo fallocate -l 8G /swapfile"
ssh netcup "sudo chmod 600 /swapfile"
ssh netcup "sudo mkswap /swapfile"
ssh netcup "sudo swapon /swapfile"
```
---
## 📊 Performance Expectations
### Text Generation:
- **Local (Llama3-70b)**: 2-10 seconds
- **Local (Mistral-7b)**: 1-3 seconds
- **RunPod (fallback)**: 3-8 seconds
- **Cost**: $0.00 (local) or $0.001-0.01 (RunPod)
### Image Generation:
- **Local SD CPU (low priority)**: 30-60 seconds
- **RunPod GPU (high priority)**: 3-10 seconds
- **Cost**: $0.00 (local) or $0.02 (RunPod)
### Video Generation:
- **RunPod Wan2.1**: 30-90 seconds
- **Cost**: ~$0.50 per video
### Expected Monthly Costs:
**Light Usage (100 requests/day):**
- 70 text (local): $0
- 20 images (15 local + 5 RunPod): $0.10
- 10 videos: $5.00
- **Total: ~$5-10/month**
**Medium Usage (500 requests/day):**
- 350 text (local): $0
- 100 images (60 local + 40 RunPod): $0.80
- 50 videos: $25.00
- **Total: ~$25-35/month**
**Heavy Usage (2000 requests/day):**
- 1400 text (local): $0
- 400 images (200 local + 200 RunPod): $4.00
- 200 videos: $100.00
- **Total: ~$100-120/month**
Compare to persistent GPU pod: $200-300/month regardless of usage!
---
## 🎯 Next Steps
1. ✅ Deploy AI Orchestrator on Netcup RS 8000
2. ✅ Setup local AI models (Ollama, SD)
3. ✅ Configure RunPod endpoints
4. ✅ Test all AI services
5. 📋 Setup monitoring and alerts
6. 📋 Configure DNS for ai-api.jeffemmett.com
7. 📋 Setup SSL with Let's Encrypt
8. 📋 Migrate canvas-website to Netcup
9. 📋 Monitor costs and optimize routing
10. 📋 Decommission DigitalOcean droplets
---
## 📚 Additional Resources
- **Migration Plan**: See `NETCUP_MIGRATION_PLAN.md`
- **RunPod Setup**: See `RUNPOD_SETUP.md`
- **Test Guide**: See `TEST_RUNPOD_AI.md`
- **API Documentation**: http://159.195.32.209:8000/docs
- **Monitoring**: http://159.195.32.209:3001 (Grafana)
---
## 💡 Tips for Cost Optimization
1. **Prefer low priority for batch jobs**: Use `priority: "low"` for non-urgent tasks
2. **Use local models first**: 70-80% of workload can run locally for $0
3. **Monitor queue depth**: Auto-scales to RunPod when local is backed up
4. **Set cost alerts**: Get notified if daily costs exceed threshold
5. **Review cost breakdown weekly**: Identify optimization opportunities
6. **Batch similar requests**: Process multiple items together
7. **Cache results**: Store and reuse common queries
---
**Ready to deploy?** Start with Step 1 and follow the guide! 🚀

372
AI_SERVICES_SUMMARY.md Normal file
View File

@ -0,0 +1,372 @@
# AI Services Setup - Complete Summary
## ✅ What We've Built
You now have a **complete, production-ready AI orchestration system** that intelligently routes between your Netcup RS 8000 (local CPU - FREE) and RunPod (serverless GPU - pay-per-use).
---
## 📦 Files Created/Modified
### New Files:
1. **`NETCUP_MIGRATION_PLAN.md`** - Complete migration plan from DigitalOcean to Netcup
2. **`AI_SERVICES_DEPLOYMENT_GUIDE.md`** - Step-by-step deployment and testing guide
3. **`src/lib/aiOrchestrator.ts`** - AI Orchestrator client library
4. **`src/shapes/VideoGenShapeUtil.tsx`** - Video generation shape (Wan2.1)
5. **`src/tools/VideoGenTool.ts`** - Video generation tool
### Modified Files:
1. **`src/shapes/ImageGenShapeUtil.tsx`** - Disabled mock mode (line 13: `USE_MOCK_API = false`)
2. **`.env.example`** - Added AI Orchestrator and RunPod configuration
### Existing Files (Already Working):
- `src/lib/runpodApi.ts` - RunPod API client for transcription
- `src/utils/llmUtils.ts` - Enhanced LLM utilities with RunPod support
- `src/hooks/useWhisperTranscriptionSimple.ts` - WhisperX transcription
- `RUNPOD_SETUP.md` - RunPod setup documentation
- `TEST_RUNPOD_AI.md` - Testing documentation
---
## 🎯 Features & Capabilities
### 1. Text Generation (LLM)
- ✅ Smart routing to local Ollama (FREE)
- ✅ Fallback to RunPod if needed
- ✅ Works with: Prompt shapes, arrow LLM actions, command palette
- ✅ Models: Llama3-70b, CodeLlama-34b, Mistral-7b, etc.
- 💰 **Cost: $0** (99% of requests use local CPU)
### 2. Image Generation
- ✅ Priority-based routing:
- Low priority → Local SD CPU (slow but FREE)
- High priority → RunPod GPU (fast, $0.02)
- ✅ Auto-scaling based on queue depth
- ✅ ImageGenShapeUtil and ImageGenTool
- ✅ Mock mode **DISABLED** - ready for production
- 💰 **Cost: $0-0.02** per image
### 3. Video Generation (NEW!)
- ✅ Wan2.1 I2V 14B 720p model on RunPod
- ✅ VideoGenShapeUtil with video player
- ✅ VideoGenTool for canvas
- ✅ Download generated videos
- ✅ Configurable duration (1-10 seconds)
- 💰 **Cost: ~$0.50** per video
### 4. Voice Transcription
- ✅ WhisperX on RunPod (primary)
- ✅ Automatic fallback to local Whisper
- ✅ TranscriptionShapeUtil
- 💰 **Cost: $0.01-0.05** per transcription
---
## 🏗️ Architecture
```
User Request
│
▼
AI Orchestrator (RS 8000)
│
├─── Text/Code ───────▶ Local Ollama (FREE)
│
├─── Images (low) ────▶ Local SD CPU (FREE, slow)
│
├─── Images (high) ───▶ RunPod GPU ($0.02, fast)
│
└─── Video ───────────▶ RunPod GPU ($0.50)
```
### Smart Routing Benefits:
- **70-80% of workload runs for FREE** (local CPU)
- **No idle GPU costs** (serverless = pay only when generating)
- **Auto-scaling** (queue-based, handles spikes)
- **Cost tracking** (per job, per user, per day/month)
- **Graceful fallback** (local → RunPod → error)
---
## 💰 Cost Analysis
### Before (DigitalOcean + Persistent GPU):
- Main Droplet: $18-36/mo
- AI Droplet: $36/mo
- RunPod persistent pods: $100-200/mo
- **Total: $154-272/mo**
### After (Netcup RS 8000 + Serverless GPU):
- RS 8000 G12 Pro: €55.57/mo (~$60/mo)
- RunPod serverless: $30-60/mo (70% reduction)
- **Total: $90-120/mo**
### Savings:
- **Monthly: $64-152**
- **Annual: $768-1,824**
### Plus You Get:
- 10x CPU cores (20 vs 2)
- 32x RAM (64GB vs 2GB)
- 25x storage (3TB vs 120GB)
- Better EU latency (Germany)
---
## 📋 Quick Start Checklist
### Phase 1: Deploy AI Orchestrator (1-2 hours)
- [ ] SSH into Netcup RS 8000: `ssh netcup`
- [ ] Create directory: `/opt/ai-orchestrator`
- [ ] Deploy docker-compose stack (see NETCUP_MIGRATION_PLAN.md Phase 2)
- [ ] Configure environment variables (.env)
- [ ] Start services: `docker-compose up -d`
- [ ] Verify: `curl http://localhost:8000/health`
### Phase 2: Setup Local AI Models (2-4 hours)
- [ ] Download Ollama models (Llama3-70b, CodeLlama-34b)
- [ ] Download Stable Diffusion 2.1 weights
- [ ] Download Wan2.1 model weights (optional, runs on RunPod)
- [ ] Test Ollama: `docker exec ai-ollama ollama run llama3:70b "Hello"`
### Phase 3: Configure RunPod Endpoints (30 min)
- [ ] Create text generation endpoint (optional)
- [ ] Create image generation endpoint (SDXL)
- [ ] Create video generation endpoint (Wan2.1)
- [ ] Copy endpoint IDs
- [ ] Update .env with endpoint IDs
- [ ] Restart services: `docker-compose restart`
### Phase 4: Configure canvas-website (15 min)
- [ ] Create `.env.local` with AI Orchestrator URL
- [ ] Add RunPod API keys (fallback)
- [ ] Install dependencies: `npm install`
- [ ] Register VideoGenShapeUtil and VideoGenTool (see deployment guide)
- [ ] Build: `npm run build`
- [ ] Start: `npm run dev`
### Phase 5: Test Everything (1 hour)
- [ ] Test AI Orchestrator health check
- [ ] Test text generation (local Ollama)
- [ ] Test image generation (low priority - local)
- [ ] Test image generation (high priority - RunPod)
- [ ] Test video generation (RunPod Wan2.1)
- [ ] Test voice transcription (WhisperX)
- [ ] Check cost tracking dashboard
- [ ] Monitor queue status
### Phase 6: Production Deployment (2-4 hours)
- [ ] Setup nginx reverse proxy
- [ ] Configure DNS: ai-api.jeffemmett.com → 159.195.32.209
- [ ] Setup SSL with Let's Encrypt
- [ ] Deploy canvas-website to RS 8000
- [ ] Setup monitoring dashboards (Grafana)
- [ ] Configure cost alerts
- [ ] Test from production domain
---
## 🧪 Testing Commands
### Test AI Orchestrator:
```bash
# Health check
curl http://159.195.32.209:8000/health
# Text generation
curl -X POST http://159.195.32.209:8000/generate/text \
-H "Content-Type: application/json" \
-d '{"prompt":"Hello world in Python","priority":"normal"}'
# Image generation (low priority)
curl -X POST http://159.195.32.209:8000/generate/image \
-H "Content-Type: application/json" \
-d '{"prompt":"A beautiful sunset","priority":"low"}'
# Video generation
curl -X POST http://159.195.32.209:8000/generate/video \
-H "Content-Type: application/json" \
-d '{"prompt":"A cat walking","duration":3}'
# Queue status
curl http://159.195.32.209:8000/queue/status
# Costs
curl http://159.195.32.209:3000/api/costs/summary
```
---
## 📊 Monitoring Dashboards
Access your monitoring at:
- **API Docs**: http://159.195.32.209:8000/docs
- **Queue Status**: http://159.195.32.209:8000/queue/status
- **Cost Tracking**: http://159.195.32.209:3000/api/costs/summary
- **Grafana**: http://159.195.32.209:3001 (login: admin/admin)
- **Prometheus**: http://159.195.32.209:9090
---
## 🔧 Configuration Files
### Environment Variables (.env.local):
```bash
# AI Orchestrator (Primary)
VITE_AI_ORCHESTRATOR_URL=http://159.195.32.209:8000
# RunPod (Fallback)
VITE_RUNPOD_API_KEY=your_api_key
VITE_RUNPOD_TEXT_ENDPOINT_ID=xxx
VITE_RUNPOD_IMAGE_ENDPOINT_ID=xxx
VITE_RUNPOD_VIDEO_ENDPOINT_ID=xxx
```
### AI Orchestrator (.env on RS 8000):
```bash
# PostgreSQL
POSTGRES_PASSWORD=generated_password
# RunPod
RUNPOD_API_KEY=your_api_key
RUNPOD_TEXT_ENDPOINT_ID=xxx
RUNPOD_IMAGE_ENDPOINT_ID=xxx
RUNPOD_VIDEO_ENDPOINT_ID=xxx
# Monitoring
GRAFANA_PASSWORD=generated_password
COST_ALERT_THRESHOLD=100
```
---
## 🐛 Common Issues & Solutions
### 1. "AI Orchestrator not available"
```bash
# Check if running
ssh netcup "cd /opt/ai-orchestrator && docker-compose ps"
# Restart
ssh netcup "cd /opt/ai-orchestrator && docker-compose restart"
# Check logs
ssh netcup "cd /opt/ai-orchestrator && docker-compose logs -f router"
```
### 2. "Image generation fails"
- Check RunPod endpoint configuration
- Verify endpoint returns: `{"output": {"image": "url"}}`
- Test endpoint directly in RunPod console
### 3. "Video generation timeout"
- Normal processing time: 30-90 seconds
- Check RunPod GPU availability (cold start can add 30s)
- Verify Wan2.1 endpoint is deployed correctly
### 4. "High costs"
```bash
# Check cost breakdown
curl http://159.195.32.209:3000/api/costs/summary
# Adjust routing to prefer local more
# Edit /opt/ai-orchestrator/services/router/main.py
# Increase queue_depth threshold from 10 to 20+
```
---
## 📚 Documentation Index
1. **NETCUP_MIGRATION_PLAN.md** - Complete migration guide (8 phases)
2. **AI_SERVICES_DEPLOYMENT_GUIDE.md** - Deployment and testing guide
3. **AI_SERVICES_SUMMARY.md** - This file (quick reference)
4. **RUNPOD_SETUP.md** - RunPod WhisperX setup
5. **TEST_RUNPOD_AI.md** - Testing guide for RunPod integration
---
## 🎯 Next Actions
**Immediate (Today):**
1. Review the migration plan (NETCUP_MIGRATION_PLAN.md)
2. Verify SSH access to Netcup RS 8000
3. Get RunPod API keys and endpoint IDs
**This Week:**
1. Deploy AI Orchestrator on Netcup (Phase 2)
2. Download local AI models (Phase 3)
3. Configure RunPod endpoints
4. Test basic functionality
**Next Week:**
1. Full testing of all AI services
2. Deploy canvas-website to Netcup
3. Setup monitoring and alerts
4. Configure DNS and SSL
**Future:**
1. Migrate remaining services from DigitalOcean
2. Decommission DigitalOcean droplets
3. Optimize costs based on usage patterns
4. Scale workers based on demand
---
## 💡 Pro Tips
1. **Start small**: Deploy text generation first, then images, then video
2. **Monitor costs daily**: Use the cost dashboard to track spending
3. **Use low priority for batch jobs**: Save 100% on images that aren't urgent
4. **Cache common results**: Store and reuse frequent queries
5. **Set cost alerts**: Get email when daily costs exceed threshold
6. **Test locally first**: Use mock API during development
7. **Review queue depths**: Optimize routing thresholds based on your usage
---
## 🚀 Expected Performance
### Text Generation:
- **Latency**: 2-10s (local), 3-8s (RunPod)
- **Throughput**: 10-20 requests/min (local)
- **Cost**: $0 (local), $0.001-0.01 (RunPod)
### Image Generation:
- **Latency**: 30-60s (local low), 3-10s (RunPod high)
- **Throughput**: 1-2 images/min (local), 6-10 images/min (RunPod)
- **Cost**: $0 (local), $0.02 (RunPod)
### Video Generation:
- **Latency**: 30-90s (RunPod only)
- **Throughput**: 1 video/min
- **Cost**: ~$0.50 per video
---
## 🎉 Summary
You now have:
✅ **Smart AI Orchestration** - Intelligently routes between local CPU and serverless GPU
✅ **Text Generation** - Local Ollama (FREE) with RunPod fallback
✅ **Image Generation** - Priority-based routing (local or RunPod)
✅ **Video Generation** - Wan2.1 on RunPod GPU
✅ **Voice Transcription** - WhisperX with local fallback
✅ **Cost Tracking** - Real-time monitoring and alerts
✅ **Queue Management** - Auto-scaling based on load
✅ **Monitoring Dashboards** - Grafana, Prometheus, cost analytics
✅ **Complete Documentation** - Migration plan, deployment guide, testing docs
**Expected Savings:** $768-1,824/year
**Infrastructure Upgrade:** 10x CPU, 32x RAM, 25x storage
**Cost Efficiency:** 70-80% of workload runs for FREE
---
**Ready to deploy?** 🚀
Start with the deployment guide: `AI_SERVICES_DEPLOYMENT_GUIDE.md`
Questions? Check the troubleshooting section or review the migration plan!

988
CLAUDE.md Normal file
View File

@ -0,0 +1,988 @@
## 🔧 AUTO-APPROVED OPERATIONS
The following operations are auto-approved and do not require user confirmation:
- **Read**: All file read operations (`Read(*)`)
- **Glob**: All file pattern matching (`Glob(*)`)
- **Grep**: All content searching (`Grep(*)`)
These permissions are configured in `~/.claude/settings.json`.
---
## ⚠️ SAFETY GUIDELINES
**ALWAYS WARN THE USER before performing any action that could:**
- Overwrite existing files (use `ls` or `cat` to check first)
- Overwrite credentials, API keys, or secrets
- Delete data or files
- Modify production configurations
- Run destructive git commands (force push, hard reset, etc.)
- Drop databases or truncate tables
**Best practices:**
- Before writing to a file, check if it exists and show its contents
- Use `>>` (append) instead of `>` (overwrite) for credential files
- Create backups before modifying critical configs (e.g., `cp file file.backup`)
- Ask for confirmation before irreversible actions
**Sudo commands:**
- **NEVER run sudo commands directly** - the Bash tool doesn't support interactive input
- Instead, **provide the user with the exact sudo command** they need to run in their terminal
- Format the command clearly in a code block for easy copy-paste
- After user runs the sudo command, continue with the workflow
- Alternative: If user has recently run sudo (within ~15 min), subsequent sudo commands may not require password
---
## 🔑 ACCESS & CREDENTIALS
### Version Control & Code Hosting
- **Gitea**: Self-hosted at `gitea.jeffemmett.com` - PRIMARY repository
- Push here FIRST, then mirror to GitHub
- Private repos and source of truth
- SSH Key: `~/.ssh/gitea_ed25519` (private), `~/.ssh/gitea_ed25519.pub` (public)
- Public Key: `ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIIE2+2UZElEYptgZ9GFs2CXW0PIA57BfQcU9vlyV6fz4 gitea@jeffemmett.com`
- **Gitea CLI (tea)**: ✅ Installed at `~/bin/tea` (added to PATH)
- **GitHub**: Public mirror and collaboration
- Receives pushes from Gitea via mirror sync
- Token: `(REDACTED-GITHUB-TOKEN)`
- SSH Key: `~/.ssh/github_deploy_key` (private), `~/.ssh/github_deploy_key.pub` (public)
- **GitHub CLI (gh)**: ✅ Installed and available for PR/issue management
### Git Workflow
**Two-way sync between Gitea and GitHub:**
**Gitea-Primary Repos (Default):**
1. Develop locally in `/home/jeffe/Github/`
2. Commit and push to Gitea first
3. Gitea automatically mirrors TO GitHub (built-in push mirror)
4. GitHub used for public collaboration and visibility
**GitHub-Primary Repos (Mirror Repos):**
For repos where GitHub is source of truth (v0.dev exports, client collabs):
1. Push to GitHub
2. Deploy webhook pulls from GitHub and deploys
3. Webhook triggers Gitea to sync FROM GitHub
### 🔀 DEV BRANCH WORKFLOW (MANDATORY)
**CRITICAL: All development work on canvas-website (and other active projects) MUST use a dev branch.**
#### Branch Strategy
```
main (production)
└── dev (integration/staging)
└── feature/* (optional feature branches)
```
#### Development Rules
1. **ALWAYS work on the `dev` branch** for new features and changes:
```bash
cd /home/jeffe/Github/canvas-website
git checkout dev
git pull origin dev
```
2. **After completing a feature**, push to dev:
```bash
git add .
git commit -m "feat: description of changes"
git push origin dev
```
3. **Update backlog task** immediately after pushing:
```bash
backlog task edit <task-id> --status "Done" --append-notes "Pushed to dev branch"
```
4. **NEVER push directly to main** - main is for tested, verified features only
5. **Merge dev → main manually** when features are verified working:
```bash
git checkout main
git pull origin main
git merge dev
git push origin main
git checkout dev # Return to dev for continued work
```
#### Complete Feature Deployment Checklist
- [ ] Work on `dev` branch (not main)
- [ ] Test locally before committing
- [ ] Commit with descriptive message
- [ ] Push to `dev` branch on Gitea
- [ ] Update backlog task status to "Done"
- [ ] Add notes to backlog task about what was implemented
- [ ] (Later) When verified working: merge dev → main manually
#### Why This Matters
- **Protects production**: main branch always has known-working code
- **Enables testing**: dev branch can be deployed to staging for verification
- **Clean history**: main only gets complete, tested features
- **Easy rollback**: if dev breaks, main is still stable
### Server Infrastructure
- **Netcup RS 8000 G12 Pro**: Primary application & AI server
- IP: `159.195.32.209`
- 20 cores, 64GB RAM, 3TB storage
- Hosts local AI models (Ollama, Stable Diffusion)
- All websites and apps deployed here in Docker containers
- Location: Germany (low latency EU)
- SSH Key (local): `~/.ssh/netcup_ed25519` (private), `~/.ssh/netcup_ed25519.pub` (public)
- Public Key: `ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIKmp4A2klKv/YIB1C6JAsb2UzvlzzE+0EcJ0jtkyFuhO netcup-rs8000@jeffemmett.com`
- SSH Access: `ssh netcup`
- **SSH Keys ON the server** (for git operations):
- Gitea: `~/.ssh/gitea_ed25519` → `ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIIE2+2UZElEYptgZ9GFs2CXW0PIA57BfQcU9vlyV6fz4 gitea@jeffemmett.com`
- GitHub: `~/.ssh/github_ed25519` → `ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIC6xXNICy0HXnqHO+U7+y7ui+pZBGe0bm0iRMS23pR1E github-deploy@netcup-rs8000`
- **RunPod**: GPU burst capacity for AI workloads
- Host: `ssh.runpod.io`
- Serverless GPU pods (pay-per-use)
- Used for: SDXL/SD3, video generation, training
- Smart routing from RS 8000 orchestrator
- SSH Key: `~/.ssh/runpod_ed25519` (private), `~/.ssh/runpod_ed25519.pub` (public)
- Public Key: `ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIAC7NYjI0U/2ChGaZBBWP7gKt/V12Ts6FgatinJOQ8JG runpod@jeffemmett.com`
- SSH Access: `ssh runpod`
- **API Key**: `(REDACTED-RUNPOD-KEY)`
- **CLI Config**: `~/.runpod/config.toml`
- **Serverless Endpoints**:
- Image (SD): `tzf1j3sc3zufsy` (Automatic1111)
- Video (Wan2.2): `4jql4l7l0yw0f3`
- Text (vLLM): `03g5hz3hlo8gr2`
- Whisper: `lrtisuv8ixbtub`
- ComfyUI: `5zurj845tbf8he`
### API Keys & Services
**IMPORTANT**: All API keys and tokens are stored securely on the Netcup server. Never store credentials locally.
- Access credentials via: `ssh netcup "cat ~/.cloudflare-credentials.env"` or `ssh netcup "cat ~/.porkbun_credentials"`
- All API operations should be performed FROM the Netcup server, not locally
#### Credential Files on Netcup (`/root/`)
| File | Contents |
|------|----------|
| `~/.cloudflare-credentials.env` | Cloudflare API tokens, account ID, tunnel token |
| `~/.cloudflare_credentials` | Legacy/DNS token |
| `~/.porkbun_credentials` | Porkbun API key and secret |
| `~/.v0_credentials` | V0.dev API key |
#### Cloudflare
- **Account ID**: `0e7b3338d5278ed1b148e6456b940913`
- **Tokens stored on Netcup** - source `~/.cloudflare-credentials.env`:
- `CLOUDFLARE_API_TOKEN` - Zone read, Worker:read/edit, R2:read/edit
- `CLOUDFLARE_TUNNEL_TOKEN` - Tunnel management
- `CLOUDFLARE_ZONE_TOKEN` - Zone:Edit, DNS:Edit (for adding domains)
#### Porkbun (Domain Registrar)
- **Credentials stored on Netcup** - source `~/.porkbun_credentials`:
- `PORKBUN_API_KEY` and `PORKBUN_SECRET_KEY`
- **API Endpoint**: `https://api-ipv4.porkbun.com/api/json/v3/`
- **API Docs**: https://porkbun.com/api/json/v3/documentation
- **Important**: JSON must have `secretapikey` before `apikey` in requests
- **Capabilities**: Update nameservers, get auth codes for transfers, manage DNS
- **Note**: Each domain must have "API Access" enabled individually in Porkbun dashboard
#### Domain Onboarding Workflow (Porkbun → Cloudflare)
Run these commands FROM Netcup (`ssh netcup`):
1. Add domain to Cloudflare (creates zone, returns nameservers)
2. Update nameservers at Porkbun to point to Cloudflare
3. Add CNAME record pointing to Cloudflare tunnel
4. Add hostname to tunnel config and restart cloudflared
5. Domain is live through the tunnel!
#### V0.dev (AI UI Generation)
- **Credentials stored on Netcup** - source `~/.v0_credentials`:
- `V0_API_KEY` - Platform API access
- **API Key**: `v1:5AwJbit4j9rhGcAKPU4XlVWs:05vyCcJLiWRVQW7Xu4u5E03G`
- **SDK**: `npm install v0-sdk` (use `v0` CLI for adding components)
- **Docs**: https://v0.app/docs/v0-platform-api
- **Capabilities**:
- List/create/update/delete projects
- Manage chats and versions
- Download generated code
- Create deployments
- Manage environment variables
- **Limitations**: GitHub-only for git integration (no Gitea/GitLab support)
- **Usage**:
```javascript
const { v0 } = require('v0-sdk');
// Uses V0_API_KEY env var automatically
const projects = await v0.projects.find();
const chats = await v0.chats.find();
```
#### Other Services
- **HuggingFace**: CLI access available for model downloads
- **RunPod**: API access for serverless GPU orchestration (see Server Infrastructure above)
### Dev Ops Stack & Principles
- **Platform**: Linux WSL2 (Ubuntu on Windows) for development
- **Working Directory**: `/home/jeffe/Github`
- **Container Strategy**:
- ALL repos should be Dockerized
- Optimized containers for production deployment
- Docker Compose for multi-service orchestration
- **Process Management**: PM2 available for Node.js services
- **Version Control**: Git configured with GitHub + Gitea mirrors
- **Package Managers**: npm/pnpm/yarn available
### 🚀 Traefik Reverse Proxy (Central Routing)
All HTTP services on Netcup RS 8000 route through Traefik for automatic service discovery.
**Architecture:**
```
Internet → Cloudflare Tunnel → Traefik (:80/:443) → Docker Services
│
├── gitea.jeffemmett.com → gitea:3000
├── mycofi.earth → mycofi:3000
├── games.jeffemmett.com → games:80
└── [auto-discovered via Docker labels]
```
**Location:** `/root/traefik/` on Netcup RS 8000
**Adding a New Service:**
```yaml
# In your docker-compose.yml, add these labels:
services:
myapp:
image: myapp:latest
labels:
- "traefik.enable=true"
- "traefik.http.routers.myapp.rule=Host(`myapp.jeffemmett.com`)"
- "traefik.http.services.myapp.loadbalancer.server.port=3000"
networks:
- traefik-public
networks:
traefik-public:
external: true
```
**Traefik Dashboard:** `http://159.195.32.209:8888` (internal only)
**SSH Git Access:**
- SSH goes direct (not through Traefik): `git.jeffemmett.com:223` → `159.195.32.209:223`
- Web UI goes through Traefik: `gitea.jeffemmett.com` → Traefik → gitea:3000
### ☁️ Cloudflare Tunnel Configuration
**Location:** `/root/cloudflared/` on Netcup RS 8000
The tunnel uses a token-based configuration managed via Cloudflare Zero Trust Dashboard.
All public hostnames should point to `http://localhost:80` (Traefik), which routes based on Host header.
**Managed hostnames:**
- `gitea.jeffemmett.com` → Traefik → Gitea
- `photos.jeffemmett.com` → Traefik → Immich
- `movies.jeffemmett.com` → Traefik → Jellyfin
- `search.jeffemmett.com` → Traefik → Semantic Search
- `mycofi.earth` → Traefik → MycoFi
- `games.jeffemmett.com` → Traefik → Games Platform
- `decolonizeti.me` → Traefik → Decolonize Time
**Tunnel ID:** `a838e9dc-0af5-4212-8af2-6864eb15e1b5`
**Tunnel CNAME Target:** `a838e9dc-0af5-4212-8af2-6864eb15e1b5.cfargotunnel.com`
**To deploy a new website/service:**
1. **Dockerize the project** with Traefik labels in `docker-compose.yml`:
```yaml
services:
myapp:
build: .
labels:
- "traefik.enable=true"
- "traefik.http.routers.myapp.rule=Host(`mydomain.com`) || Host(`www.mydomain.com`)"
- "traefik.http.services.myapp.loadbalancer.server.port=3000"
networks:
- traefik-public
networks:
traefik-public:
external: true
```
2. **Deploy to Netcup:**
```bash
ssh netcup "cd /opt/websites && git clone <repo-url>"
ssh netcup "cd /opt/websites/<project> && docker compose up -d --build"
```
3. **Add hostname to tunnel config** (`/root/cloudflared/config.yml`):
```yaml
- hostname: mydomain.com
service: http://localhost:80
- hostname: www.mydomain.com
service: http://localhost:80
```
Then restart: `ssh netcup "docker restart cloudflared"`
4. **Configure DNS in Cloudflare dashboard** (CRITICAL - prevents 525 SSL errors):
- Go to Cloudflare Dashboard → select domain → DNS → Records
- Delete any existing A/AAAA records for `@` and `www`
- Add CNAME records:
| Type | Name | Target | Proxy |
|------|------|--------|-------|
| CNAME | `@` | `a838e9dc-0af5-4212-8af2-6864eb15e1b5.cfargotunnel.com` | Proxied ✓ |
| CNAME | `www` | `a838e9dc-0af5-4212-8af2-6864eb15e1b5.cfargotunnel.com` | Proxied ✓ |
**API Credentials** (on Netcup at `~/.cloudflare*`):
- `CLOUDFLARE_API_TOKEN` - Zone read access only
- `CLOUDFLARE_TUNNEL_TOKEN` - Tunnel management only
- See **API Keys & Services** section above for Domain Management Token (required for DNS automation)
### 🔄 Auto-Deploy Webhook System
**Location:** `/opt/deploy-webhook/` on Netcup RS 8000
**Endpoint:** `https://deploy.jeffemmett.com/deploy/<repo-name>`
**Secret:** `gitea-deploy-secret-2025`
Pushes to Gitea automatically trigger rebuilds. The webhook receiver:
1. Validates HMAC signature from Gitea
2. Runs `git pull && docker compose up -d --build`
3. Returns build status
**Adding a new repo to auto-deploy:**
1. Add entry to `/opt/deploy-webhook/webhook.py` REPOS dict
2. Restart: `ssh netcup "cd /opt/deploy-webhook && docker compose up -d --build"`
3. Add Gitea webhook:
```bash
curl -X POST "https://gitea.jeffemmett.com/api/v1/repos/jeffemmett/<repo>/hooks" \
-H "Authorization: token <gitea-token>" \
-H "Content-Type: application/json" \
-d '{"type":"gitea","active":true,"events":["push"],"config":{"url":"https://deploy.jeffemmett.com/deploy/<repo>","content_type":"json","secret":"gitea-deploy-secret-2025"}}'
```
**Currently auto-deploying:**
- `decolonize-time-website` → /opt/websites/decolonize-time-website
- `mycofi-earth-website` → /opt/websites/mycofi-earth-website
- `games-platform` → /opt/apps/games-platform
### 🔐 SSH Keys Quick Reference
**Local keys** (in `~/.ssh/` on your laptop):
| Service | Private Key | Public Key | Purpose |
|---------|-------------|------------|---------|
| **Gitea** | `gitea_ed25519` | `gitea_ed25519.pub` | Primary git repository |
| **GitHub** | `github_deploy_key` | `github_deploy_key.pub` | Public mirror sync |
| **Netcup RS 8000** | `netcup_ed25519` | `netcup_ed25519.pub` | Primary server SSH |
| **RunPod** | `runpod_ed25519` | `runpod_ed25519.pub` | GPU pods SSH |
| **Default** | `id_ed25519` | `id_ed25519.pub` | General purpose/legacy |
**Server-side keys** (in `/root/.ssh/` on Netcup RS 8000):
| Service | Key File | Purpose |
|---------|----------|---------|
| **Gitea** | `gitea_ed25519` | Server pulls from Gitea repos |
| **GitHub** | `github_ed25519` | Server pulls from GitHub (mirror repos) |
**SSH Config**: `~/.ssh/config` contains all host configurations
**Quick Access**:
- `ssh netcup` - Connect to Netcup RS 8000
- `ssh runpod` - Connect to RunPod
- `ssh gitea.jeffemmett.com` - Git operations
---
## 🤖 AI ORCHESTRATION ARCHITECTURE
### Smart Routing Strategy
All AI requests go through intelligent orchestration layer on RS 8000:
**Routing Logic:**
- **Text/Code (70-80% of workload)**: Always local RS 8000 CPU (Ollama) → FREE
- **Images - Low Priority**: RS 8000 CPU (SD 1.5/2.1) → FREE but slow (~60s)
- **Images - High Priority**: RunPod GPU (SDXL/SD3) → $0.02/image, fast
- **Video Generation**: Always RunPod GPU → $0.50/video (only option)
- **Training/Fine-tuning**: RunPod GPU on-demand
**Queue System:**
- Redis-based queues: text, image, code, video
- Priority-based routing (low/normal/high)
- Worker pools scale based on load
- Cost tracking per job, per user
**Cost Optimization:**
- Target: $90-120/mo (vs $136-236/mo current)
- Savings: $552-1,392/year
- 70-80% of workload FREE (local CPU)
- GPU only when needed (serverless = no idle costs)
### Deployment Architecture
```
RS 8000 G12 Pro (Netcup)
├── Cloudflare Tunnel (secure ingress)
├── Traefik Reverse Proxy (auto-discovery)
│ └── Routes to all services via Docker labels
├── Core Services
│ ├── Gitea (git hosting) - gitea.jeffemmett.com
│ └── Other internal tools
├── AI Services
│ ├── Ollama (text/code models)
│ ├── Stable Diffusion (CPU fallback)
│ └── Smart Router API (FastAPI)
├── Queue Infrastructure
│ ├── Redis (job queues)
│ └── PostgreSQL (job history/analytics)
├── Monitoring
│ ├── Prometheus (metrics)
│ ├── Grafana (dashboards)
│ └── Cost tracking API
└── Application Hosting
├── All websites (Dockerized + Traefik labels)
├── All apps (Dockerized + Traefik labels)
└── Backend services (Dockerized)
RunPod Serverless (GPU Burst)
├── SDXL/SD3 endpoints
├── Video generation (Wan2.1)
└── Training/fine-tuning jobs
```
### Integration Pattern for Projects
All projects use unified AI client SDK:
```python
from orchestrator_client import AIOrchestrator
ai = AIOrchestrator("http://rs8000-ip:8000")
# Automatically routes based on priority & model
result = await ai.generate_text(prompt, priority="low") # → FREE CPU
result = await ai.generate_image(prompt, priority="high") # → RunPod GPU
```
---
## 💰 GPU COST ANALYSIS & MIGRATION PLAN
### Current Infrastructure Costs (Monthly)
| Service | Type | Cost | Notes |
|---------|------|------|-------|
| Netcup RS 8000 G12 Pro | Fixed | ~€45 | 20 cores, 64GB RAM, 3TB (CPU-only) |
| RunPod Serverless | Variable | $50-100 | Pay-per-use GPU (images, video) |
| DigitalOcean Droplets | Fixed | ~$48 | ⚠️ DEPRECATED - migrate ASAP |
| **Current Total** | | **~$140-190/mo** | |
### GPU Provider Comparison
#### Netcup vGPU (NEW - Early Access, Ends July 7, 2025)
| Plan | GPU | VRAM | vCores | RAM | Storage | Price/mo | Price/hr equiv |
|------|-----|------|--------|-----|---------|----------|----------------|
| RS 2000 vGPU 7 | H200 | 7 GB dedicated | 8 | 16 GB DDR5 | 512 GB NVMe | €137.31 (~$150) | $0.21/hr |
| RS 4000 vGPU 14 | H200 | 14 GB dedicated | 12 | 32 GB DDR5 | 1 TB NVMe | €261.39 (~$285) | $0.40/hr |
**Pros:**
- NVIDIA H200 (latest gen, better than H100 for inference)
- Dedicated VRAM (no noisy neighbors)
- Germany location (EU data sovereignty, low latency to RS 8000)
- Fixed monthly cost = predictable budgeting
- 24/7 availability, no cold starts
**Cons:**
- Pay even when idle
- Limited to 7GB or 14GB VRAM options
- Early access = limited availability
#### RunPod Serverless (Current)
| GPU | VRAM | Price/hr | Typical Use |
|-----|------|----------|-------------|
| RTX 4090 | 24 GB | ~$0.44/hr | SDXL, medium models |
| A100 40GB | 40 GB | ~$1.14/hr | Large models, training |
| H100 80GB | 80 GB | ~$2.49/hr | Largest models |
**Current Endpoint Costs:**
- Image (SD/SDXL): ~$0.02/image (~2s compute)
- Video (Wan2.2): ~$0.50/video (~60s compute)
- Text (vLLM): ~$0.001/request
- Whisper: ~$0.01/minute audio
**Pros:**
- Zero idle costs
- Unlimited burst capacity
- Wide GPU selection (up to 80GB VRAM)
- Pay only for actual compute
**Cons:**
- Cold start delays (10-30s first request)
- Variable availability during peak times
- Per-request costs add up at scale
### Break-even Analysis
**When does Netcup vGPU become cheaper than RunPod?**
| Scenario | RunPod Cost | Netcup RS 2000 vGPU 7 | Netcup RS 4000 vGPU 14 |
|----------|-------------|----------------------|------------------------|
| 1,000 images/mo | $20 | $150 ❌ | $285 ❌ |
| 5,000 images/mo | $100 | $150 ❌ | $285 ❌ |
| **7,500 images/mo** | **$150** | **$150 ✅** | $285 ❌ |
| 10,000 images/mo | $200 | $150 ✅ | $285 ❌ |
| **14,250 images/mo** | **$285** | $150 ✅ | **$285 ✅** |
| 100 videos/mo | $50 | $150 ❌ | $285 ❌ |
| **300 videos/mo** | **$150** | **$150 ✅** | $285 ❌ |
| 500 videos/mo | $250 | $150 ✅ | $285 ❌ |
**Recommendation by Usage Pattern:**
| Monthly Usage | Best Option | Est. Cost |
|---------------|-------------|-----------|
| < 5,000 images OR < 250 videos | RunPod Serverless | $50-100 |
| 5,000-10,000 images OR 250-500 videos | **Netcup RS 2000 vGPU 7** | $150 fixed |
| > 10,000 images OR > 500 videos + training | **Netcup RS 4000 vGPU 14** | $285 fixed |
| Unpredictable/bursty workloads | RunPod Serverless | Variable |
### Migration Strategy
#### Phase 1: Immediate (Before July 7, 2025)
**Decision Point: Secure Netcup vGPU Early Access?**
- [ ] Monitor actual GPU usage for 2-4 weeks
- [ ] Calculate average monthly image/video generation
- [ ] If consistently > 5,000 images/mo → Consider RS 2000 vGPU 7
- [ ] If consistently > 10,000 images/mo → Consider RS 4000 vGPU 14
- [ ] **ACTION**: Redeem early access code if usage justifies fixed GPU
#### Phase 2: Hybrid Architecture (If vGPU Acquired)
```
RS 8000 G12 Pro (CPU - Current)
├── Ollama (text/code) → FREE
├── SD 1.5/2.1 CPU fallback → FREE
└── Orchestrator API
Netcup vGPU Server (NEW - If purchased)
├── Primary GPU workloads
├── SDXL/SD3 generation
├── Video generation (Wan2.1 I2V)
├── Model inference (14B params with 14GB VRAM)
└── Connected via internal netcup network (low latency)
RunPod Serverless (Burst Only)
├── Overflow capacity
├── Models requiring > 14GB VRAM
├── Training/fine-tuning jobs
└── Geographic distribution needs
```
#### Phase 3: Cost Optimization Targets
| Scenario | Current | With vGPU Migration | Savings |
|----------|---------|---------------------|---------|
| Low usage | $140/mo | $95/mo (RS8000 + minimal RunPod) | $540/yr |
| Medium usage | $190/mo | $195/mo (RS8000 + vGPU 7) | Break-even |
| High usage | $250/mo | $195/mo (RS8000 + vGPU 7) | $660/yr |
| Very high usage | $350/mo | $330/mo (RS8000 + vGPU 14) | $240/yr |
### Model VRAM Requirements Reference
| Model | VRAM Needed | Fits vGPU 7? | Fits vGPU 14? |
|-------|-------------|--------------|---------------|
| SD 1.5 | ~4 GB | ✅ | ✅ |
| SD 2.1 | ~5 GB | ✅ | ✅ |
| SDXL | ~7 GB | ⚠️ Tight | ✅ |
| SD3 Medium | ~8 GB | ❌ | ✅ |
| Wan2.1 I2V 14B | ~12 GB | ❌ | ✅ |
| Wan2.1 T2V 14B | ~14 GB | ❌ | ⚠️ Tight |
| Flux.1 Dev | ~12 GB | ❌ | ✅ |
| LLaMA 3 8B (Q4) | ~6 GB | ✅ | ✅ |
| LLaMA 3 70B (Q4) | ~40 GB | ❌ | ❌ (RunPod) |
### Decision Framework
```
┌─────────────────────────────────────────────────────────┐
│ GPU WORKLOAD DECISION TREE │
├─────────────────────────────────────────────────────────┤
│ │
│ Is usage predictable and consistent? │
│ ├── YES → Is monthly GPU spend > $150? │
│ │ ├── YES → Netcup vGPU (fixed cost wins) │
│ │ └── NO → RunPod Serverless (no idle cost) │
│ └── NO → RunPod Serverless (pay for what you use) │
│ │
│ Does model require > 14GB VRAM? │
│ ├── YES → RunPod (A100/H100 on-demand) │
│ └── NO → Netcup vGPU or RS 8000 CPU │
│ │
│ Is low latency critical? │
│ ├── YES → Netcup vGPU (same datacenter as RS 8000) │
│ └── NO → RunPod Serverless (acceptable for batch) │
│ │
└─────────────────────────────────────────────────────────┘
```
### Monitoring & Review Schedule
- **Weekly**: Review RunPod spend dashboard
- **Monthly**: Calculate total GPU costs, compare to vGPU break-even
- **Quarterly**: Re-evaluate architecture, consider plan changes
- **Annually**: Full infrastructure cost audit
### Action Items
- [ ] **URGENT**: Decide on Netcup vGPU early access before July 7, 2025
- [ ] Set up GPU usage tracking in orchestrator
- [ ] Create Grafana dashboard for cost monitoring
- [ ] Test Wan2.1 I2V 14B model on vGPU 14 (if acquired)
- [ ] Document migration runbook for vGPU setup
- [ ] Complete DigitalOcean deprecation (separate from GPU decision)
---
## 📁 PROJECT PORTFOLIO STRUCTURE
### Repository Organization
- **Location**: `/home/jeffe/Github/`
- **Primary Flow**: Gitea (source of truth) → GitHub (public mirror)
- **Containerization**: ALL repos must be Dockerized with optimized production containers
### 🎯 MAIN PROJECT: canvas-website
**Location**: `/home/jeffe/Github/canvas-website`
**Description**: Collaborative canvas deployment - the integration hub where all tools come together
- Tldraw-based collaborative canvas platform
- Integrates Hyperindex, rSpace, MycoFi, and other tools
- Real-time collaboration features
- Deployed on RS 8000 in Docker
- Uses AI orchestrator for intelligent features
### Project Categories
**AI & Infrastructure:**
- AI Orchestrator (smart routing between RS 8000 & RunPod)
- Model hosting & fine-tuning pipelines
- Cost optimization & monitoring dashboards
**Web Applications & Sites:**
- **canvas-website**: Main collaborative canvas (integration hub)
- All deployed in Docker containers on RS 8000
- Cloudflare Workers for edge functions (Hyperindex)
- Static sites + dynamic backends containerized
**Supporting Projects:**
- **Hyperindex**: Tldraw canvas integration (Cloudflare stack) - integrates into canvas-website
- **rSpace**: Real-time collaboration platform - integrates into canvas-website
- **MycoFi**: DeFi/Web3 project - integrates into canvas-website
- **Canvas-related tools**: Knowledge graph & visualization components
### Deployment Strategy
1. **Development**: Local WSL2 environment (`/home/jeffe/Github/`)
2. **Version Control**: Push to Gitea FIRST → Auto-mirror to GitHub
3. **Containerization**: Build optimized Docker images with Traefik labels
4. **Deployment**: Deploy to RS 8000 via Docker Compose (join `traefik-public` network)
5. **Routing**: Traefik auto-discovers service via labels, no config changes needed
6. **DNS**: Add hostname to Cloudflare tunnel (if new domain) or it just works (existing domains)
7. **AI Integration**: Connect to local orchestrator API
8. **Monitoring**: Grafana dashboards for all services
### Infrastructure Philosophy
- **Self-hosted first**: Own your infrastructure (RS 8000 + Gitea)
- **Cloud for edge cases**: Cloudflare (edge), RunPod (GPU burst)
- **Cost-optimized**: Local CPU for 70-80% of workload
- **Dockerized everything**: Reproducible, scalable, maintainable
- **Smart orchestration**: Right compute for the right job
---
- can you make sure you are runing the hf download for a non deprecated version? After that, you can proceed with Image-to-Video 14B 720p (RECOMMENDED)
huggingface-cli download Wan-AI/Wan2.1-I2V-14B-720P \
--include "*.safetensors" \
--local-dir models/diffusion_models/wan2.1_i2v_14b
## 🕸️ HYPERINDEX PROJECT - TOP PRIORITY
**Location:** `/home/jeffe/Github/hyperindex-system/`
When user is ready to work on the hyperindexing system:
1. Reference `HYPERINDEX_PROJECT.md` for complete architecture and implementation details
2. Follow `HYPERINDEX_TODO.md` for step-by-step checklist
3. Start with Phase 1 (Database & Core Types), then proceed sequentially through Phase 5
4. This is a tldraw canvas integration project using Cloudflare Workers, D1, R2, and Durable Objects
5. Creates a "living, mycelial network" of web discoveries that spawn on the canvas in real-time
---
## 📋 BACKLOG.MD - UNIFIED TASK MANAGEMENT
**All projects use Backlog.md for task tracking.** Tasks are managed as markdown files and can be viewed at `backlog.jeffemmett.com` for a unified cross-project view.
### MCP Integration
Backlog.md is integrated via MCP server. Available tools:
- `backlog.task_create` - Create new tasks
- `backlog.task_list` - List tasks with filters
- `backlog.task_update` - Update task status/details
- `backlog.task_view` - View task details
- `backlog.search` - Search across tasks, docs, decisions
### Task Lifecycle Workflow
**CRITICAL: Claude agents MUST follow this workflow for ALL development tasks:**
#### 1. Task Discovery (Before Starting Work)
```bash
# Check if task already exists
backlog search "<task description>" --plain
# List current tasks
backlog task list --plain
```
#### 2. Task Creation (If Not Exists)
```bash
# Create task with full details
backlog task create "Task Title" \
--desc "Detailed description" \
--priority high \
--status "To Do"
```
#### 3. Starting Work (Move to In Progress)
```bash
# Update status when starting
backlog task edit <task-id> --status "In Progress"
```
#### 4. During Development (Update Notes)
```bash
# Append progress notes
backlog task edit <task-id> --append-notes "Completed X, working on Y"
# Update acceptance criteria
backlog task edit <task-id> --check-ac 1
```
#### 5. Completion (Move to Done)
```bash
# Mark complete when finished
backlog task edit <task-id> --status "Done"
```
### Project Initialization
When starting work in a new repository that doesn't have backlog:
```bash
cd /path/to/repo
backlog init "Project Name" --integration-mode mcp --defaults
```
This creates the `backlog/` directory structure:
```
backlog/
├── config.yml # Project configuration
├── tasks/ # Active tasks
├── completed/ # Finished tasks
├── drafts/ # Draft tasks
├── docs/ # Project documentation
├── decisions/ # Architecture decision records
└── archive/ # Archived tasks
```
### Task File Format
Tasks are markdown files with YAML frontmatter:
```yaml
---
id: task-001
title: Feature implementation
status: In Progress
assignee: [@claude]
created_date: '2025-12-03 14:30'
labels: [feature, backend]
priority: high
dependencies: [task-002]
---
## Description
What needs to be done...
## Plan
1. Step one
2. Step two
## Acceptance Criteria
- [ ] Criterion 1
- [x] Criterion 2 (completed)
## Notes
Progress updates go here...
```
### Cross-Project Aggregation (backlog.jeffemmett.com)
**Architecture:**
```
┌─────────────────────────────────────────────────────────────┐
│ backlog.jeffemmett.com │
│ (Unified Kanban Dashboard) │
├─────────────────────────────────────────────────────────────┤
│ │
│ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ │
│ │ canvas-web │ │ hyperindex │ │ mycofi │ ... │
│ │ (purple) │ │ (green) │ │ (blue) │ │
│ └──────┬──────┘ └──────┬──────┘ └──────┬──────┘ │
│ │ │ │ │
│ └────────────────┴────────────────┘ │
│ │ │
│ ┌───────────┴───────────┐ │
│ │ Aggregation API │ │
│ │ (polls all projects) │ │
│ └───────────────────────┘ │
│ │
└─────────────────────────────────────────────────────────────┘
Data Sources:
├── Local: /home/jeffe/Github/*/backlog/
└── Remote: ssh netcup "ls /opt/*/backlog/"
```
**Color Coding by Project:**
| Project | Color | Location |
|---------|-------|----------|
| canvas-website | Purple | Local + Netcup |
| hyperindex-system | Green | Local |
| mycofi-earth | Blue | Local + Netcup |
| decolonize-time | Orange | Local + Netcup |
| ai-orchestrator | Red | Netcup |
**Aggregation Service** (to be deployed on Netcup):
- Polls all project `backlog/tasks/` directories
- Serves unified JSON API at `api.backlog.jeffemmett.com`
- Web UI at `backlog.jeffemmett.com` shows combined Kanban
- Real-time updates via WebSocket
- Filter by project, status, priority, assignee
### Agent Behavior Requirements
**When Claude starts working on ANY task:**
1. **Check for existing backlog** in the repo:
```bash
ls backlog/config.yml 2>/dev/null || echo "Backlog not initialized"
```
2. **If backlog exists**, search for related tasks:
```bash
backlog search "<relevant keywords>" --plain
```
3. **Create or update task** before writing code:
```bash
# If new task needed:
backlog task create "Task title" --status "In Progress"
# If task exists:
backlog task edit <id> --status "In Progress"
```
4. **Update task on completion**:
```bash
backlog task edit <id> --status "Done" --append-notes "Implementation complete"
```
5. **Never leave tasks in "In Progress"** when stopping work - either complete them or add notes explaining blockers.
### Viewing Tasks
**Terminal Kanban Board:**
```bash
backlog board
```
**Web Interface (single project):**
```bash
backlog browser --port 6420
```
**Unified View (all projects):**
Visit `backlog.jeffemmett.com` (served from Netcup)
### Backlog CLI Quick Reference
#### Task Operations
| Action | Command |
|--------|---------|
| View task | `backlog task 42 --plain` |
| List tasks | `backlog task list --plain` |
| Search tasks | `backlog search "topic" --plain` |
| Filter by status | `backlog task list -s "In Progress" --plain` |
| Create task | `backlog task create "Title" -d "Description" --ac "Criterion 1"` |
| Edit task | `backlog task edit 42 -t "New Title" -s "In Progress"` |
| Assign task | `backlog task edit 42 -a @claude` |
#### Acceptance Criteria Management
| Action | Command |
|--------|---------|
| Add AC | `backlog task edit 42 --ac "New criterion"` |
| Check AC #1 | `backlog task edit 42 --check-ac 1` |
| Check multiple | `backlog task edit 42 --check-ac 1 --check-ac 2` |
| Uncheck AC | `backlog task edit 42 --uncheck-ac 1` |
| Remove AC | `backlog task edit 42 --remove-ac 2` |
#### Multi-line Input (Description/Plan/Notes)
The CLI preserves input literally. Use shell-specific syntax for real newlines:
```bash
# Bash/Zsh (ANSI-C quoting)
backlog task edit 42 --notes $'Line1\nLine2\nLine3'
backlog task edit 42 --plan $'1. Step one\n2. Step two'
# POSIX portable
backlog task edit 42 --notes "$(printf 'Line1\nLine2')"
# Append notes progressively
backlog task edit 42 --append-notes $'- Completed X\n- Working on Y'
```
#### Definition of Done (DoD)
A task is **Done** only when ALL of these are complete:
**Via CLI:**
1. All acceptance criteria checked: `--check-ac <index>` for each
2. Implementation notes added: `--notes "..."` or `--append-notes "..."`
3. Status set to Done: `-s Done`
**Via Code/Testing:**
4. Tests pass (run test suite and linting)
5. Documentation updated if needed
6. Code self-reviewed
7. No regressions
**NEVER mark a task as Done without completing ALL items above.**
### Configuration Reference
---
## 🔧 TROUBLESHOOTING
### tmux "server exited unexpectedly"
This error occurs when a stale socket file exists from a crashed tmux server.
**Fix:**
```bash
rm -f /tmp/tmux-$(id -u)/default
```
Then start a new session normally with `tmux` or `tmux new -s <name>`.
---
Default `backlog/config.yml`:
```yaml
project_name: "Project Name"
default_status: "To Do"
statuses: ["To Do", "In Progress", "Done"]
labels: []
milestones: []
date_format: yyyy-mm-dd
max_column_width: 20
auto_open_browser: true
default_port: 6420
remote_operations: true
auto_commit: true
zero_padded_ids: 3
bypass_git_hooks: false
check_active_branches: true
active_branch_days: 60
```

1
DATA_CONVERSION_GUIDE.md
View File

@ -73,7 +73,6 @@ Custom shape types are preserved:
- ObsNote - ObsNote
- Holon - Holon
- FathomMeetingsBrowser - FathomMeetingsBrowser
- FathomTranscript
- HolonBrowser - HolonBrowser
- LocationShare - LocationShare
- ObsidianBrowser - ObsidianBrowser

54
Dockerfile Normal file
View File

@ -0,0 +1,54 @@
# Canvas Website Dockerfile
# Builds Vite frontend and serves with nginx
# Backend (sync) still uses Cloudflare Workers
# Build stage
FROM node:20-alpine AS build
WORKDIR /app
# Install dependencies
COPY package*.json ./
RUN npm ci --legacy-peer-deps
# Copy source
COPY . .
# Build args for environment
ARG VITE_TLDRAW_WORKER_URL=https://jeffemmett-canvas.jeffemmett.workers.dev
ARG VITE_DAILY_API_KEY
ARG VITE_RUNPOD_API_KEY
ARG VITE_RUNPOD_IMAGE_ENDPOINT_ID
ARG VITE_RUNPOD_VIDEO_ENDPOINT_ID
ARG VITE_RUNPOD_TEXT_ENDPOINT_ID
ARG VITE_RUNPOD_WHISPER_ENDPOINT_ID
# Set environment for build
ENV VITE_TLDRAW_WORKER_URL=$VITE_TLDRAW_WORKER_URL
ENV VITE_DAILY_API_KEY=$VITE_DAILY_API_KEY
ENV VITE_RUNPOD_API_KEY=$VITE_RUNPOD_API_KEY
ENV VITE_RUNPOD_IMAGE_ENDPOINT_ID=$VITE_RUNPOD_IMAGE_ENDPOINT_ID
ENV VITE_RUNPOD_VIDEO_ENDPOINT_ID=$VITE_RUNPOD_VIDEO_ENDPOINT_ID
ENV VITE_RUNPOD_TEXT_ENDPOINT_ID=$VITE_RUNPOD_TEXT_ENDPOINT_ID
ENV VITE_RUNPOD_WHISPER_ENDPOINT_ID=$VITE_RUNPOD_WHISPER_ENDPOINT_ID
# Build the app
RUN npm run build
# Production stage
FROM nginx:alpine AS production
WORKDIR /usr/share/nginx/html
# Remove default nginx static assets
RUN rm -rf ./*
# Copy built assets from build stage
COPY --from=build /app/dist .
# Copy nginx config
COPY nginx.conf /etc/nginx/conf.d/default.conf
# Expose port
EXPOSE 80
# Start nginx
CMD ["nginx", "-g", "daemon off;"]

232
MULTMUX_INTEGRATION.md Normal file
View File

@ -0,0 +1,232 @@
# mulTmux Integration
mulTmux is now integrated into the canvas-website project as a collaborative terminal tool. This allows multiple developers to work together in the same terminal session.
## Installation
From the root of the canvas-website project:
```bash
# Install all dependencies including mulTmux packages
npm run multmux:install
# Build mulTmux packages
npm run multmux:build
```
## Available Commands
All commands are run from the **root** of the canvas-website project:
| Command | Description |
|---------|-------------|
| `npm run multmux:install` | Install mulTmux dependencies |
| `npm run multmux:build` | Build server and CLI packages |
| `npm run multmux:dev:server` | Run server in development mode |
| `npm run multmux:dev:cli` | Run CLI in development mode |
| `npm run multmux:start` | Start the production server |
## Quick Start
### 1. Build mulTmux
```bash
npm run multmux:build
```
### 2. Start the Server Locally (for testing)
```bash
npm run multmux:start
```
Server will be available at:
- HTTP API: `http://localhost:3000`
- WebSocket: `ws://localhost:3001`
### 3. Install CLI Globally
```bash
cd multmux/packages/cli
npm link
```
Now you can use the `multmux` command anywhere!
### 4. Create a Session
```bash
# Local testing
multmux create my-session
# Or specify your AI server (when deployed)
multmux create my-session --server http://your-ai-server:3000
```
### 5. Join from Another Terminal
```bash
multmux join <token-from-above> --server ws://your-ai-server:3001
```
## Deploying to AI Server
### Option 1: Using the Deploy Script
```bash
cd multmux
./infrastructure/deploy.sh
```
This will:
- Install system dependencies (tmux, Node.js)
- Build the project
- Set up PM2 for process management
- Start the server
### Option 2: Manual Deployment
1. **SSH to your AI server**
```bash
ssh your-ai-server
```
2. **Clone or copy the project**
```bash
git clone <your-repo>
cd canvas-website
git checkout mulTmux-webtree
```
3. **Install and build**
```bash
npm install
npm run multmux:build
```
4. **Start with PM2**
```bash
cd multmux
npm install -g pm2
pm2 start packages/server/dist/index.js --name multmux-server
pm2 save
pm2 startup
```
## Project Structure
```
canvas-website/
├── multmux/
│ ├── packages/
│ │ ├── server/ # Backend (Node.js + tmux)
│ │ └── cli/ # Command-line client
│ ├── infrastructure/
│ │ ├── deploy.sh # Auto-deployment script
│ │ └── nginx.conf # Reverse proxy config
│ └── README.md # Full documentation
├── package.json # Now includes workspace config
└── MULTMUX_INTEGRATION.md # This file
```
## Usage Examples
### Collaborative Coding Session
```bash
# Developer 1: Create session in project directory
cd /path/to/project
multmux create coding-session --repo $(pwd)
# Developer 2: Join and start coding together
multmux join <token>
# Both can now type in the same terminal!
```
### Debugging Together
```bash
# Create a session for debugging
multmux create debug-auth-issue
# Share token with teammate
# Both can run commands, check logs, etc.
```
### List Active Sessions
```bash
multmux list
```
## Configuration
### Environment Variables
You can customize ports by setting environment variables:
```bash
export PORT=3000 # HTTP API port
export WS_PORT=3001 # WebSocket port
```
### Token Expiration
Default: 60 minutes. To change, edit `/home/jeffe/Github/canvas-website/multmux/packages/server/src/managers/TokenManager.ts:11`
### Session Cleanup
Sessions auto-cleanup when all users disconnect. To change this behavior, edit `/home/jeffe/Github/canvas-website/multmux/packages/server/src/managers/SessionManager.ts:64`
## Troubleshooting
### "Command not found: multmux"
Run `npm link` from the CLI package:
```bash
cd multmux/packages/cli
npm link
```
### "Connection refused"
1. Check server is running:
```bash
pm2 status
```
2. Check ports are available:
```bash
netstat -tlnp | grep -E '3000|3001'
```
3. Check logs:
```bash
pm2 logs multmux-server
```
### Token Expired
Generate a new token:
```bash
curl -X POST http://localhost:3000/api/sessions/<session-id>/tokens \
-H "Content-Type: application/json" \
-d '{"expiresInMinutes": 60}'
```
## Security Notes
- Tokens expire after 60 minutes
- Sessions are isolated per tmux instance
- All input is validated on the server
- Use nginx + SSL for production deployments
## Next Steps
1. **Test locally first**: Run `npm run multmux:start` and try creating/joining sessions
2. **Deploy to AI server**: Use `./infrastructure/deploy.sh`
3. **Set up nginx**: Copy config from `infrastructure/nginx.conf` for SSL/reverse proxy
4. **Share with team**: Send them tokens to collaborate!
For full documentation, see `multmux/README.md`.

1519
NETCUP_MIGRATION_PLAN.md Normal file
View File

File diff suppressed because it is too large Load Diff

267
QUICK_START.md Normal file
View File

@ -0,0 +1,267 @@
# Quick Start Guide - AI Services Setup
**Get your AI orchestration running in under 30 minutes!**
---
## 🎯 Goal
Deploy a smart AI orchestration layer that saves you $768-1,824/year by routing 70-80% of workload to your Netcup RS 8000 (FREE) and only using RunPod GPU when needed.
---
## ⚡ 30-Minute Quick Start
### Step 1: Verify Access (2 min)
```bash
# Test SSH to Netcup RS 8000
ssh netcup "hostname && docker --version"
# Expected output:
# vXXXXXX.netcup.net
# Docker version 24.0.x
```
✅ **Success?** Continue to Step 2
❌ **Failed?** Setup SSH key or contact Netcup support
### Step 2: Deploy AI Orchestrator (10 min)
```bash
# Create directory structure
ssh netcup << 'EOF'
mkdir -p /opt/ai-orchestrator/{services/{router,workers,monitor},configs,data}
cd /opt/ai-orchestrator
EOF
# Deploy minimal stack (text generation only for quick start)
ssh netcup "cat > /opt/ai-orchestrator/docker-compose.yml" << 'EOF'
version: '3.8'
services:
redis:
image: redis:7-alpine
ports: ["6379:6379"]
volumes: ["./data/redis:/data"]
command: redis-server --appendonly yes
ollama:
image: ollama/ollama:latest
ports: ["11434:11434"]
volumes: ["/data/models/ollama:/root/.ollama"]
EOF
# Start services
ssh netcup "cd /opt/ai-orchestrator && docker-compose up -d"
# Verify
ssh netcup "docker ps"
```
### Step 3: Download AI Model (5 min)
```bash
# Pull Llama 3 8B (smaller, faster for testing)
ssh netcup "docker exec ollama ollama pull llama3:8b"
# Test it
ssh netcup "docker exec ollama ollama run llama3:8b 'Hello, world!'"
```
Expected output: A friendly AI response!
### Step 4: Test from Your Machine (3 min)
```bash
# Get Netcup IP
NETCUP_IP="159.195.32.209"
# Test Ollama directly
curl -X POST http://$NETCUP_IP:11434/api/generate \
-H "Content-Type: application/json" \
-d '{
"model": "llama3:8b",
"prompt": "Write hello world in Python",
"stream": false
}'
```
Expected: Python code response!
### Step 5: Configure canvas-website (5 min)
```bash
cd /home/jeffe/Github/canvas-website-branch-worktrees/add-runpod-AI-API
# Create minimal .env.local
cat > .env.local << 'EOF'
# Ollama direct access (for quick testing)
VITE_OLLAMA_URL=http://159.195.32.209:11434
# Your existing vars...
VITE_GOOGLE_CLIENT_ID=your_google_client_id
VITE_TLDRAW_WORKER_URL=your_worker_url
EOF
# Install and start
npm install
npm run dev
```
### Step 6: Test in Browser (5 min)
1. Open http://localhost:5173 (or your dev port)
2. Create a Prompt shape or use LLM command
3. Type: "Write a hello world program"
4. Submit
5. Verify: Response appears using your local Ollama!
**🎉 Success!** You're now running AI locally for FREE!
---
## 🚀 Next: Full Setup (Optional)
Once quick start works, deploy the full stack:
### Option A: Full AI Orchestrator (1 hour)
Follow: `AI_SERVICES_DEPLOYMENT_GUIDE.md` Phase 2-3
Adds:
- Smart routing layer
- Image generation (local SD + RunPod)
- Video generation (RunPod Wan2.1)
- Cost tracking
- Monitoring dashboards
### Option B: Just Add Image Generation (30 min)
```bash
# Add Stable Diffusion CPU to docker-compose.yml
ssh netcup "cat >> /opt/ai-orchestrator/docker-compose.yml" << 'EOF'
stable-diffusion:
image: ghcr.io/stablecog/sc-worker:latest
ports: ["7860:7860"]
volumes: ["/data/models/stable-diffusion:/models"]
environment:
USE_CPU: "true"
EOF
ssh netcup "cd /opt/ai-orchestrator && docker-compose up -d"
```
### Option C: Full Migration (4-5 weeks)
Follow: `NETCUP_MIGRATION_PLAN.md` for complete DigitalOcean → Netcup migration
---
## 🐛 Quick Troubleshooting
### "Connection refused to 159.195.32.209:11434"
```bash
# Check if firewall blocking
ssh netcup "sudo ufw status"
ssh netcup "sudo ufw allow 11434/tcp"
ssh netcup "sudo ufw allow 8000/tcp" # For AI orchestrator later
```
### "docker: command not found"
```bash
# Install Docker
ssh netcup << 'EOF'
curl -fsSL https://get.docker.com -o get-docker.sh
sudo sh get-docker.sh
sudo usermod -aG docker $USER
EOF
# Reconnect and retry
ssh netcup "docker --version"
```
### "Ollama model not found"
```bash
# List installed models
ssh netcup "docker exec ollama ollama list"
# If empty, pull model
ssh netcup "docker exec ollama ollama pull llama3:8b"
```
### "AI response very slow (>30s)"
```bash
# Check if downloading model for first time
ssh netcup "docker exec ollama ollama list"
# Use smaller model for testing
ssh netcup "docker exec ollama ollama pull mistral:7b"
```
---
## 💡 Quick Tips
1. **Start with 8B model**: Faster responses, good for testing
2. **Use localhost for dev**: Point directly to Ollama URL
3. **Deploy orchestrator later**: Once basic setup works
4. **Monitor resources**: `ssh netcup htop` to check CPU/RAM
5. **Test locally first**: Verify before adding RunPod costs
---
## 📋 Checklist
- [ ] SSH access to Netcup works
- [ ] Docker installed and running
- [ ] Redis and Ollama containers running
- [ ] Llama3 model downloaded
- [ ] Test curl request works
- [ ] canvas-website .env.local configured
- [ ] Browser test successful
**All checked?** You're ready! 🎉
---
## 🎯 Next Steps
Choose your path:
**Path 1: Keep it Simple**
- Use Ollama directly for text generation
- Add user API keys in canvas settings for images
- Deploy full orchestrator later
**Path 2: Deploy Full Stack**
- Follow `AI_SERVICES_DEPLOYMENT_GUIDE.md`
- Setup image + video generation
- Enable cost tracking and monitoring
**Path 3: Full Migration**
- Follow `NETCUP_MIGRATION_PLAN.md`
- Migrate all services from DigitalOcean
- Setup production infrastructure
---
## 📚 Reference Docs
- **This Guide**: Quick 30-min setup
- **AI_SERVICES_SUMMARY.md**: Complete feature overview
- **AI_SERVICES_DEPLOYMENT_GUIDE.md**: Full deployment (all services)
- **NETCUP_MIGRATION_PLAN.md**: Complete migration plan (8 phases)
- **RUNPOD_SETUP.md**: RunPod WhisperX setup
- **TEST_RUNPOD_AI.md**: Testing guide
---
**Questions?** Check `AI_SERVICES_SUMMARY.md` or deployment guide!
**Ready for full setup?** Continue to `AI_SERVICES_DEPLOYMENT_GUIDE.md`! 🚀

255
RUNPOD_SETUP.md Normal file
View File

@ -0,0 +1,255 @@
# RunPod WhisperX Integration Setup
This guide explains how to set up and use the RunPod WhisperX endpoint for transcription in the canvas website.
## Overview
The transcription system can now use a hosted WhisperX endpoint on RunPod instead of running the Whisper model locally in the browser. This provides:
- Better accuracy with WhisperX's advanced features
- Faster processing (no model download needed)
- Reduced client-side resource usage
- Support for longer audio files
## Prerequisites
1. A RunPod account with an active WhisperX endpoint
2. Your RunPod API key
3. Your RunPod endpoint ID
## Configuration
### Environment Variables
Add the following environment variables to your `.env.local` file (or your deployment environment):
```bash
# RunPod Configuration
VITE_RUNPOD_API_KEY=your_runpod_api_key_here
VITE_RUNPOD_ENDPOINT_ID=your_endpoint_id_here
```
Or if using Next.js:
```bash
NEXT_PUBLIC_RUNPOD_API_KEY=your_runpod_api_key_here
NEXT_PUBLIC_RUNPOD_ENDPOINT_ID=your_endpoint_id_here
```
### Getting Your RunPod Credentials
1. **API Key**:
- Go to [RunPod Settings](https://www.runpod.io/console/user/settings)
- Navigate to API Keys section
- Create a new API key or copy an existing one
2. **Endpoint ID**:
- Go to [RunPod Serverless Endpoints](https://www.runpod.io/console/serverless)
- Find your WhisperX endpoint
- Copy the endpoint ID from the URL or endpoint details
- Example: If your endpoint URL is `https://api.runpod.ai/v2/lrtisuv8ixbtub/run`, then `lrtisuv8ixbtub` is your endpoint ID
## Usage
### Automatic Detection
The transcription hook automatically detects if RunPod is configured and uses it instead of the local Whisper model. No code changes are needed!
### Manual Override
If you want to explicitly control which transcription method to use:
```typescript
import { useWhisperTranscription } from '@/hooks/useWhisperTranscriptionSimple'
const {
isRecording,
transcript,
startRecording,
stopRecording
} = useWhisperTranscription({
useRunPod: true, // Force RunPod usage
language: 'en',
onTranscriptUpdate: (text) => {
console.log('New transcript:', text)
}
})
```
Or to force local model:
```typescript
useWhisperTranscription({
useRunPod: false, // Force local Whisper model
// ... other options
})
```
## API Format
The integration sends audio data to your RunPod endpoint in the following format:
```json
{
"input": {
"audio": "base64_encoded_audio_data",
"audio_format": "audio/wav",
"language": "en",
"task": "transcribe"
}
}
```
### Expected Response Format
The endpoint should return one of these formats:
**Direct Response:**
```json
{
"output": {
"text": "Transcribed text here"
}
}
```
**Or with segments:**
```json
{
"output": {
"segments": [
{
"start": 0.0,
"end": 2.5,
"text": "Transcribed text here"
}
]
}
}
```
**Async Job Pattern:**
```json
{
"id": "job-id-123",
"status": "IN_QUEUE"
}
```
The integration automatically handles async jobs by polling the status endpoint until completion.
## Customizing the API Request
If your WhisperX endpoint expects a different request format, you can modify `src/lib/runpodApi.ts`:
```typescript
// In transcribeWithRunPod function
const requestBody = {
input: {
// Adjust these fields based on your endpoint
audio: audioBase64,
// Add or modify fields as needed
}
}
```
## Troubleshooting
### "RunPod API key or endpoint ID not configured"
- Ensure environment variables are set correctly
- Restart your development server after adding environment variables
- Check that variable names match exactly (case-sensitive)
### "RunPod API error: 401"
- Verify your API key is correct
- Check that your API key has not expired
- Ensure you're using the correct API key format
### "RunPod API error: 404"
- Verify your endpoint ID is correct
- Check that your endpoint is active in the RunPod console
- Ensure the endpoint URL format matches: `https://api.runpod.ai/v2/{ENDPOINT_ID}/run`
### "No transcription text found in RunPod response"
- Check your endpoint's response format matches the expected format
- Verify your WhisperX endpoint is configured correctly
- Check the browser console for detailed error messages
### "Failed to return job results" (400 Bad Request)
This error occurs on the **server side** when your WhisperX endpoint tries to return results. This typically means:
1. **Response format mismatch**: Your endpoint's response doesn't match RunPod's expected format
- Ensure your endpoint returns: `{"output": {"text": "..."}}` or `{"output": {"segments": [...]}}`
- The response must be valid JSON
- Check your endpoint handler code to ensure it's returning the correct structure
2. **Response size limits**: The response might be too large
- Try with shorter audio files first
- Check RunPod's response size limits
3. **Timeout issues**: The endpoint might be taking too long to process
- Check your endpoint logs for processing time
- Consider optimizing your WhisperX model configuration
4. **Check endpoint handler**: Review your WhisperX endpoint's `handler.py` or equivalent:
```python
# Example correct format
def handler(event):
# ... process audio ...
return {
"output": {
"text": transcription_text
}
}
```
### Transcription not working
- Check browser console for errors
- Verify your endpoint is active and responding
- Test your endpoint directly using curl or Postman
- Ensure audio format is supported (WAV format is recommended)
- Check RunPod endpoint logs for server-side errors
## Testing Your Endpoint
You can test your RunPod endpoint directly:
```bash
curl -X POST https://api.runpod.ai/v2/YOUR_ENDPOINT_ID/run \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"input": {
"audio": "base64_audio_data_here",
"audio_format": "audio/wav",
"language": "en"
}
}'
```
## Fallback Behavior
If RunPod is not configured or fails, the system will:
1. Try to use RunPod if configured
2. Fall back to local Whisper model if RunPod fails or is not configured
3. Show error messages if both methods fail
## Performance Considerations
- **RunPod**: Better for longer audio files and higher accuracy, but requires network connection
- **Local Model**: Works offline, but requires model download and uses more client resources
## Support
For issues specific to:
- **RunPod API**: Check [RunPod Documentation](https://docs.runpod.io)
- **WhisperX**: Check your WhisperX endpoint configuration
- **Integration**: Check browser console for detailed error messages

139
TEST_RUNPOD_AI.md Normal file
View File

@ -0,0 +1,139 @@
# Testing RunPod AI Integration
This guide explains how to test the RunPod AI API integration in development.
## Quick Setup
1. **Add RunPod environment variables to `.env.local`:**
```bash
# Add these lines to your .env.local file
VITE_RUNPOD_API_KEY=your_runpod_api_key_here
VITE_RUNPOD_ENDPOINT_ID=your_endpoint_id_here
```
**Important:** Replace `your_runpod_api_key_here` and `your_endpoint_id_here` with your actual RunPod credentials.
2. **Get your RunPod credentials:**
- **API Key**: Go to [RunPod Settings](https://www.runpod.io/console/user/settings) → API Keys section
- **Endpoint ID**: Go to [RunPod Serverless Endpoints](https://www.runpod.io/console/serverless) → Find your endpoint → Copy the ID from the URL
- Example: If URL is `https://api.runpod.ai/v2/jqd16o7stu29vq/run`, then `jqd16o7stu29vq` is your endpoint ID
3. **Restart the dev server:**
```bash
npm run dev
```
## Testing the Integration
### Method 1: Using Prompt Shapes
1. Open the canvas website in your browser
2. Select the **Prompt** tool from the toolbar (or press the keyboard shortcut)
3. Click on the canvas to create a prompt shape
4. Type a prompt like "Write a hello world program in Python"
5. Press Enter or click the send button
6. The AI response should appear in the prompt shape
### Method 2: Using Arrow LLM Action
1. Create an arrow shape pointing from one shape to another
2. Add text to the arrow (this becomes the prompt)
3. Select the arrow
4. Press **Alt+G** (or use the action menu)
5. The AI will process the prompt and fill the target shape with the response
### Method 3: Using Command Palette
1. Press **Cmd+J** (Mac) or **Ctrl+J** (Windows/Linux) to open the LLM view
2. Type your prompt
3. Press Enter
4. The response should appear
## Verifying RunPod is Being Used
1. **Open browser console** (F12 or Cmd+Option+I)
2. Look for these log messages:
- `🔑 Found RunPod configuration from environment variables - using as primary AI provider`
- `🔍 Found X available AI providers: runpod (default)`
- `🔄 Attempting to use runpod API (default)...`
3. **Check Network tab:**
- Look for requests to `https://api.runpod.ai/v2/{endpointId}/run`
- The request should have `Authorization: Bearer {your_api_key}` header
## Expected Behavior
- **With RunPod configured**: RunPod will be used FIRST (priority over user API keys)
- **Without RunPod**: System will fall back to user-configured API keys (OpenAI, Anthropic, etc.)
- **If both fail**: You'll see an error message
## Troubleshooting
### "No valid API key found for any provider"
- Check that `.env.local` has the correct variable names (`VITE_RUNPOD_API_KEY` and `VITE_RUNPOD_ENDPOINT_ID`)
- Restart the dev server after adding environment variables
- Check browser console for detailed error messages
### "RunPod API error: 401"
- Verify your API key is correct
- Check that your API key hasn't expired
- Ensure you're using the correct API key format
### "RunPod API error: 404"
- Verify your endpoint ID is correct
- Check that your endpoint is active in RunPod console
- Ensure the endpoint URL format matches: `https://api.runpod.ai/v2/{ENDPOINT_ID}/run`
### RunPod not being used
- Check browser console for `🔑 Found RunPod configuration` message
- Verify environment variables are loaded (check `import.meta.env.VITE_RUNPOD_API_KEY` in console)
- Make sure you restarted the dev server after adding environment variables
## Testing Different Scenarios
### Test 1: RunPod Only (No User Keys)
1. Remove or clear any user API keys from localStorage
2. Set RunPod environment variables
3. Run an AI command
4. Should use RunPod automatically
### Test 2: RunPod Priority (With User Keys)
1. Set RunPod environment variables
2. Also configure user API keys in settings
3. Run an AI command
4. Should use RunPod FIRST, then fall back to user keys if RunPod fails
### Test 3: Fallback Behavior
1. Set RunPod environment variables with invalid credentials
2. Configure valid user API keys
3. Run an AI command
4. Should try RunPod first, fail, then use user keys
## API Request Format
The integration sends requests in this format:
```json
{
"input": {
"prompt": "Your prompt text here"
}
}
```
The system prompt and user prompt are combined into a single prompt string.
## Response Handling
The integration handles multiple response formats:
- Direct text response: `{ "output": "text" }`
- Object with text: `{ "output": { "text": "..." } }`
- Object with response: `{ "output": { "response": "..." } }`
- Async jobs: Polls until completion
## Next Steps
Once testing is successful:
1. Verify RunPod responses are working correctly
2. Test with different prompt types
3. Monitor RunPod usage and costs
4. Consider adding rate limiting if needed

341
WORKTREE_SETUP.md Normal file
View File

@ -0,0 +1,341 @@
# Git Worktree Automation Setup
This repository is configured to automatically create Git worktrees for new branches, allowing you to work on multiple branches simultaneously without switching contexts.
## What Are Worktrees?
Git worktrees allow you to have multiple working directories (copies of your repo) checked out to different branches at the same time. This means:
- No need to stash or commit work when switching branches
- Run dev servers on multiple branches simultaneously
- Compare code across branches easily
- Keep your main branch clean while working on features
## Automatic Worktree Creation
A Git hook (`.git/hooks/post-checkout`) is installed that automatically creates worktrees when you create a new branch from `main`:
```bash
# This will automatically create a worktree at ../canvas-website-feature-name
git checkout -b feature/new-feature
```
**Worktree Location Pattern:**
```
/home/jeffe/Github/
├── canvas-website/ # Main repo (main branch)
├── canvas-website-feature-name/ # Worktree for feature branch
└── canvas-website-bugfix-something/ # Worktree for bugfix branch
```
## Manual Worktree Management
Use the `worktree-manager.sh` script for manual management:
### List All Worktrees
```bash
./scripts/worktree-manager.sh list
```
### Create a New Worktree
```bash
# Creates worktree for existing branch
./scripts/worktree-manager.sh create feature/my-feature
# Or create new branch with worktree
./scripts/worktree-manager.sh create feature/new-branch
```
### Remove a Worktree
```bash
./scripts/worktree-manager.sh remove feature/old-feature
```
### Clean Up All Worktrees (Keep Main)
```bash
./scripts/worktree-manager.sh clean
```
### Show Status of All Worktrees
```bash
./scripts/worktree-manager.sh status
```
### Navigate to a Worktree
```bash
# Get worktree path
./scripts/worktree-manager.sh goto feature/my-feature
# Or use with cd
cd $(./scripts/worktree-manager.sh goto feature/my-feature)
```
### Help
```bash
./scripts/worktree-manager.sh help
```
## Workflow Examples
### Starting a New Feature
**With automatic worktree creation:**
```bash
# In main repo
cd /home/jeffe/Github/canvas-website
# Create and switch to new branch (worktree auto-created)
git checkout -b feature/terminal-tool
# Notification appears:
# 🌳 Creating worktree for branch: feature/terminal-tool
# 📁 Location: /home/jeffe/Github/canvas-website-feature-terminal-tool
# Continue working in current directory or switch to worktree
cd ../canvas-website-feature-terminal-tool
```
**Manual worktree creation:**
```bash
./scripts/worktree-manager.sh create feature/my-feature
cd $(./scripts/worktree-manager.sh goto feature/my-feature)
```
### Working on Multiple Features Simultaneously
```bash
# Terminal 1: Main repo (main branch)
cd /home/jeffe/Github/canvas-website
npm run dev # Port 5173
# Terminal 2: Feature branch 1
cd /home/jeffe/Github/canvas-website-feature-auth
npm run dev # Different port
# Terminal 3: Feature branch 2
cd /home/jeffe/Github/canvas-website-feature-ui
npm run dev # Another port
# All running simultaneously, no conflicts!
```
### Comparing Code Across Branches
```bash
# Use diff or your IDE to compare files
diff /home/jeffe/Github/canvas-website/src/App.tsx \
/home/jeffe/Github/canvas-website-feature-auth/src/App.tsx
# Or open both in VS Code
code /home/jeffe/Github/canvas-website \
/home/jeffe/Github/canvas-website-feature-auth
```
### Cleaning Up After Merging
```bash
# After merging feature/my-feature to main
cd /home/jeffe/Github/canvas-website
# Remove the worktree
./scripts/worktree-manager.sh remove feature/my-feature
# Or clean all worktrees except main
./scripts/worktree-manager.sh clean
```
## How It Works
### Post-Checkout Hook
The `.git/hooks/post-checkout` script runs automatically after `git checkout` and:
1. Detects if you're creating a new branch from `main`
2. Creates a worktree in `../canvas-website-{branch-name}`
3. Links the worktree to the new branch
4. Shows a notification with the worktree path
**Hook Behavior:**
- ✅ Creates worktree when: `git checkout -b new-branch` (from main)
- ❌ Skips creation when:
- Switching to existing branches
- Already in a worktree
- Worktree already exists for that branch
- Not branching from main/master
### Worktree Manager Script
The `scripts/worktree-manager.sh` script provides:
- User-friendly commands for worktree operations
- Colored output for better readability
- Error handling and validation
- Status reporting across all worktrees
## Git Commands with Worktrees
Most Git commands work the same way in worktrees:
```bash
# In any worktree
git status # Shows status of current worktree
git add . # Stages files in current worktree
git commit -m "..." # Commits in current branch
git push # Pushes current branch
git pull # Pulls current branch
# List all worktrees (works from any worktree)
git worktree list
# Remove a worktree (from main repo)
git worktree remove feature/branch-name
# Prune deleted worktrees
git worktree prune
```
## Important Notes
### Shared Git Directory
All worktrees share the same `.git` directory (in the main repo), which means:
- ✅ Commits, branches, and remotes are shared across all worktrees
- ✅ One `git fetch` or `git pull` in main updates all worktrees
- ⚠️ Don't delete the main repo while worktrees exist
- ⚠️ Stashes are shared (stash in one worktree, pop in another)
### Node Modules
Each worktree has its own `node_modules`:
- First time entering a worktree: run `npm install`
- Dependencies may differ across branches
- More disk space usage (one `node_modules` per worktree)
### Port Conflicts
When running dev servers in multiple worktrees:
```bash
# Main repo
npm run dev # Uses default port 5173
# In worktree, specify different port
npm run dev -- --port 5174
```
### IDE Integration
**VS Code:**
```bash
# Open specific worktree
code /home/jeffe/Github/canvas-website-feature-name
# Or open multiple worktrees as workspace
code --add /home/jeffe/Github/canvas-website \
--add /home/jeffe/Github/canvas-website-feature-name
```
## Troubleshooting
### Worktree Path Already Exists
If you see:
```
fatal: '/path/to/worktree' already exists
```
Remove the directory manually:
```bash
rm -rf /home/jeffe/Github/canvas-website-feature-name
git worktree prune
```
### Can't Delete Main Repo
If you have active worktrees, you can't delete the main repo. Clean up first:
```bash
./scripts/worktree-manager.sh clean
```
### Worktree Out of Sync
If a worktree seems out of sync:
```bash
cd /path/to/worktree
git fetch origin
git reset --hard origin/branch-name
```
### Hook Not Running
If the post-checkout hook isn't running:
```bash
# Check if it's executable
ls -la .git/hooks/post-checkout
# Make it executable if needed
chmod +x .git/hooks/post-checkout
# Test the hook manually
.git/hooks/post-checkout HEAD HEAD 1
```
## Disabling Automatic Worktrees
To disable automatic worktree creation:
```bash
# Remove or rename the hook
mv .git/hooks/post-checkout .git/hooks/post-checkout.disabled
```
To re-enable:
```bash
mv .git/hooks/post-checkout.disabled .git/hooks/post-checkout
```
## Advanced Usage
### Custom Worktree Location
Modify the `post-checkout` hook to change the worktree location:
```bash
# Edit .git/hooks/post-checkout
# Change this line:
WORKTREE_BASE=$(dirname "$REPO_ROOT")
# To (example):
WORKTREE_BASE="$HOME/worktrees"
```
### Worktree for Remote Branches
```bash
# Create worktree for remote branch
git worktree add ../canvas-website-remote-branch origin/feature-branch
# Or use the script
./scripts/worktree-manager.sh create origin/feature-branch
```
### Detached HEAD Worktree
```bash
# Create worktree at specific commit
git worktree add ../canvas-website-commit-abc123 abc123
```
## Best Practices
1. **Clean up regularly**: Remove worktrees for merged branches
2. **Name branches clearly**: Worktree names mirror branch names
3. **Run npm install**: Always run in new worktrees
4. **Check branch**: Always verify which branch you're on before committing
5. **Use status command**: Check all worktrees before major operations
## Resources
- [Git Worktree Documentation](https://git-scm.com/docs/git-worktree)
- [Git Hooks Documentation](https://git-scm.com/docs/githooks)
---
**Setup Complete!** New branches will automatically create worktrees. Use `./scripts/worktree-manager.sh help` for manual management.

29
_redirects
View File

@ -1,14 +1,25 @@
# Cloudflare Pages redirects and rewrites # Cloudflare Pages redirects and rewrites
# This file handles SPA routing and URL rewrites (replaces vercel.json rewrites) # This file handles SPA routing and URL rewrites (replaces vercel.json rewrites)
# SPA fallback - all routes should serve index.html # Specific route rewrites (matching vercel.json)
# Handle both with and without trailing slashes
/board/* /index.html 200
/board /index.html 200
/board/ /index.html 200
/inbox /index.html 200
/inbox/ /index.html 200
/contact /index.html 200
/contact/ /index.html 200
/presentations /index.html 200
/presentations/ /index.html 200
/presentations/* /index.html 200
/dashboard /index.html 200
/dashboard/ /index.html 200
/login /index.html 200
/login/ /index.html 200
/debug /index.html 200
/debug/ /index.html 200
# SPA fallback - all routes should serve index.html (must be last)
/* /index.html 200 /* /index.html 200
# Specific route rewrites (matching vercel.json)
/board/* /index.html 200
/board /index.html 200
/inbox /index.html 200
/contact /index.html 200
/presentations /index.html 200
/dashboard /index.html 200

15
backlog/config.yml Normal file
View File

@ -0,0 +1,15 @@
project_name: "Canvas Feature List"
default_status: "To Do"
statuses: ["To Do", "In Progress", "Done"]
labels: []
milestones: []
date_format: yyyy-mm-dd
max_column_width: 20
auto_open_browser: true
default_port: 6420
remote_operations: true
auto_commit: true
zero_padded_ids: 3
bypass_git_hooks: false
check_active_branches: true
active_branch_days: 60

12
backlog/tasks/task-001 - offline-local-storage.md Normal file
View File

@ -0,0 +1,12 @@
---
id: task-001
title: offline local storage
status: To Do
assignee: []
created_date: '2025-12-03 23:42'
updated_date: '2025-12-04 12:13'
labels: []
dependencies: []
---

26
backlog/tasks/task-002 - runpod-ai-api-integration.md Normal file
View File

@ -0,0 +1,26 @@
---
id: task-002
title: RunPod AI API Integration
status: Done
assignee: []
created_date: '2025-12-03'
labels: [feature, ai, integration]
priority: high
branch: add-runpod-AI-API
worktree: /home/jeffe/Github/canvas-website-branch-worktrees/add-runpod-AI-API
updated_date: '2025-12-04 13:43'
---
## Description
Integrate RunPod serverless AI API for image generation and other AI features on the canvas.
## Branch Info
- **Branch**: `add-runpod-AI-API`
- **Worktree**: `/home/jeffe/Github/canvas-website-branch-worktrees/add-runpod-AI-API`
- **Commit**: 083095c
## Acceptance Criteria
- [ ] Connect to RunPod serverless endpoints
- [ ] Implement image generation from canvas
- [ ] Handle AI responses and display on canvas
- [ ] Error handling and loading states

24
backlog/tasks/task-003 - multmux-webtree.md Normal file
View File

@ -0,0 +1,24 @@
---
id: task-003
title: MulTmux Web Integration
status: In Progress
assignee: []
created_date: '2025-12-03'
labels: [feature, terminal, integration]
priority: medium
branch: mulTmux-webtree
worktree: /home/jeffe/Github/canvas-website-branch-worktrees/mulTmux-webtree
---
## Description
Integrate MulTmux web terminal functionality into the canvas for terminal-based interactions.
## Branch Info
- **Branch**: `mulTmux-webtree`
- **Worktree**: `/home/jeffe/Github/canvas-website-branch-worktrees/mulTmux-webtree`
- **Commit**: 8ea3490
## Acceptance Criteria
- [ ] Embed terminal component in canvas
- [ ] Handle terminal I/O within canvas context
- [ ] Support multiple terminal sessions

24
backlog/tasks/task-004 - io-chip-feature.md Normal file
View File

@ -0,0 +1,24 @@
---
id: task-004
title: IO Chip Feature
status: In Progress
assignee: []
created_date: '2025-12-03'
labels: [feature, io, ui]
priority: medium
branch: feature/io-chip
worktree: /home/jeffe/Github/canvas-website-io-chip
---
## Description
Implement IO chip feature for the canvas - enabling input/output connections between canvas elements.
## Branch Info
- **Branch**: `feature/io-chip`
- **Worktree**: `/home/jeffe/Github/canvas-website-io-chip`
- **Commit**: 527462a
## Acceptance Criteria
- [ ] Create IO chip component
- [ ] Enable connections between canvas elements
- [ ] Handle data flow between connected chips

22
backlog/tasks/task-005 - automerge-crdt-sync.md Normal file
View File

@ -0,0 +1,22 @@
---
id: task-005
title: Automerge CRDT Sync
status: To Do
assignee: []
created_date: '2025-12-03'
labels: [feature, sync, collaboration]
priority: high
branch: Automerge
---
## Description
Implement Automerge CRDT-based synchronization for real-time collaborative canvas editing.
## Branch Info
- **Branch**: `Automerge`
## Acceptance Criteria
- [ ] Integrate Automerge library
- [ ] Enable real-time sync between clients
- [ ] Handle conflict resolution automatically
- [ ] Persist state across sessions

22
backlog/tasks/task-006 - stripe-integration.md Normal file
View File

@ -0,0 +1,22 @@
---
id: task-006
title: Stripe Payment Integration
status: To Do
assignee: []
created_date: '2025-12-03'
labels: [feature, payments, integration]
priority: medium
branch: stripe-integration
---
## Description
Integrate Stripe for payment processing and subscription management.
## Branch Info
- **Branch**: `stripe-integration`
## Acceptance Criteria
- [ ] Set up Stripe API connection
- [ ] Implement payment flow
- [ ] Handle subscriptions
- [ ] Add billing management UI

21
backlog/tasks/task-007 - web3-integration.md Normal file
View File

@ -0,0 +1,21 @@
---
id: task-007
title: Web3 Integration
status: To Do
assignee: []
created_date: '2025-12-03'
labels: [feature, web3, blockchain]
priority: low
branch: web3-integration
---
## Description
Integrate Web3 capabilities for blockchain-based features (wallet connect, NFT canvas elements, etc.).
## Branch Info
- **Branch**: `web3-integration`
## Acceptance Criteria
- [ ] Add wallet connection
- [ ] Enable NFT minting of canvas elements
- [ ] Blockchain-based ownership verification

22
backlog/tasks/task-008 - audio-recording.md Normal file
View File

@ -0,0 +1,22 @@
---
id: task-008
title: Audio Recording Feature
status: To Do
assignee: []
created_date: '2025-12-03'
labels: [feature, audio, media]
priority: medium
branch: audio-recording-attempt
---
## Description
Implement audio recording capability for voice notes and audio annotations on the canvas.
## Branch Info
- **Branch**: `audio-recording-attempt`
## Acceptance Criteria
- [ ] Record audio from microphone
- [ ] Save audio clips to canvas
- [ ] Playback audio annotations
- [ ] Transcription integration

22
backlog/tasks/task-009 - transcribe-webspeech.md Normal file
View File

@ -0,0 +1,22 @@
---
id: task-009
title: Web Speech API Transcription
status: To Do
assignee: []
created_date: '2025-12-03'
labels: [feature, transcription, speech]
priority: medium
branch: transcribe-webspeechAPI
---
## Description
Implement speech-to-text transcription using the Web Speech API for voice input on the canvas.
## Branch Info
- **Branch**: `transcribe-webspeechAPI`
## Acceptance Criteria
- [ ] Capture speech via Web Speech API
- [ ] Convert to text in real-time
- [ ] Display transcription on canvas
- [ ] Support multiple languages

21
backlog/tasks/task-010 - holon-integration.md Normal file
View File

@ -0,0 +1,21 @@
---
id: task-010
title: Holon Integration
status: To Do
assignee: []
created_date: '2025-12-03'
labels: [feature, holon, integration]
priority: medium
branch: holon-integration
---
## Description
Integrate Holon framework for hierarchical canvas organization and nested structures.
## Branch Info
- **Branch**: `holon-integration`
## Acceptance Criteria
- [ ] Implement holon data structure
- [ ] Enable nested canvas elements
- [ ] Support hierarchical navigation

21
backlog/tasks/task-011 - terminal-tool.md Normal file
View File

@ -0,0 +1,21 @@
---
id: task-011
title: Terminal Tool
status: To Do
assignee: []
created_date: '2025-12-03'
labels: [feature, terminal, tool]
priority: medium
branch: feature/terminal-tool
---
## Description
Add a terminal tool to the canvas toolbar for embedding terminal sessions.
## Branch Info
- **Branch**: `feature/terminal-tool`
## Acceptance Criteria
- [ ] Add terminal tool to toolbar
- [ ] Spawn terminal instances on canvas
- [ ] Handle terminal sizing and positioning

67
backlog/tasks/task-012 - Dark-Mode-Theme.md Normal file
View File

@ -0,0 +1,67 @@
---
id: task-012
title: Dark Mode Theme
status: Done
assignee: []
created_date: '2025-12-03'
updated_date: '2025-12-04 06:29'
labels:
- feature
- ui
- theme
dependencies: []
priority: medium
---
## Description
<!-- SECTION:DESCRIPTION:BEGIN -->
Implement dark mode theme support for the canvas interface.
## Branch Info
- **Branch**: `dark-mode`
<!-- SECTION:DESCRIPTION:END -->
## Acceptance Criteria
<!-- AC:BEGIN -->
- [x] #1 Create dark theme colors
- [x] #2 Add theme toggle
- [x] #3 Persist user preference
- [x] #4 System theme detection
<!-- AC:END -->
## Implementation Notes
<!-- SECTION:NOTES:BEGIN -->
## Implementation Complete (2025-12-03)
### Components Updated:
1. **Mycelial Intelligence (MI) Bar** (`src/ui/MycelialIntelligenceBar.tsx`)
- Added dark mode color palette with automatic switching based on `isDark` state
- Dark backgrounds, lighter text, adjusted shadows
- Inline code blocks use CSS class for proper dark mode styling
2. **Comprehensive CSS Dark Mode** (`src/css/style.css`)
- Added CSS variables: `--card-bg`, `--input-bg`, `--muted-text`
- Dark mode styles for: blockquotes, tables, navigation, command palette, MDXEditor, chat containers, form inputs, error/success messages
3. **UserSettingsModal** (`src/ui/UserSettingsModal.tsx`)
- Added `colors` object with dark/light mode variants
- Updated all inline styles to use theme-aware colors
4. **StandardizedToolWrapper** (`src/components/StandardizedToolWrapper.tsx`)
- Added `useIsDarkMode` hook for dark mode detection
- Updated wrapper backgrounds, shadows, borders, tags styling
5. **Markdown Tool** (`src/shapes/MarkdownShapeUtil.tsx`)
- Dark mode detection with automatic background switching
- Fixed scrollbar: vertical only, hidden when not needed
- Added toolbar minimize/expand button
### Technical Details:
- Automatic detection via `document.documentElement.classList` observer
- CSS variables for base styles that auto-switch in dark mode
- Inline style support with conditional color objects
- Comprehensive coverage of all major UI components and tools
<!-- SECTION:NOTES:END -->

44
backlog/tasks/task-013 - Markdown-Tool-UX-Improvements.md Normal file
View File

@ -0,0 +1,44 @@
---
id: task-013
title: Markdown Tool UX Improvements
status: Done
assignee: []
created_date: '2025-12-04 06:29'
updated_date: '2025-12-04 06:29'
labels:
- feature
- ui
- markdown
dependencies: []
priority: medium
---
## Description
<!-- SECTION:DESCRIPTION:BEGIN -->
Improve the Markdown tool user experience with better scrollbar behavior and collapsible toolbar.
## Changes Implemented:
- Scrollbar is now vertical only (no horizontal scrollbar)
- Scrollbar auto-hides when not needed
- Added minimize/expand button for the formatting toolbar
- Full editing area uses available space
<!-- SECTION:DESCRIPTION:END -->
## Acceptance Criteria
<!-- AC:BEGIN -->
- [x] #1 Scrollbar is vertical only
- [x] #2 Scrollbar hides when not needed
- [x] #3 Toolbar has minimize/expand toggle
- [x] #4 Full window is editing area
<!-- AC:END -->
## Implementation Notes
<!-- SECTION:NOTES:BEGIN -->
Implementation completed in `src/shapes/MarkdownShapeUtil.tsx`:
- Added `overflow-x: hidden` to content area
- Custom scrollbar styling with thin width and auto-hide
- Added toggle button in toolbar that collapses/expands formatting options
- `isToolbarMinimized` state controls toolbar visibility
<!-- SECTION:NOTES:END -->

351
backlog/tasks/task-014 - Implement-WebGPU-based-local-image-generation-to-reduce-RunPod-costs.md Normal file
View File

@ -0,0 +1,351 @@
---
id: task-014
title: Implement WebGPU-based local image generation to reduce RunPod costs
status: To Do
assignee: []
created_date: '2025-12-04 11:46'
updated_date: '2025-12-04 11:47'
labels:
- performance
- cost-optimization
- webgpu
- ai
- image-generation
dependencies: []
priority: high
---
## Description
<!-- SECTION:DESCRIPTION:BEGIN -->
Integrate WebGPU-powered browser-based image generation (SD-Turbo) to reduce RunPod API costs and eliminate cold start delays. This creates a hybrid pipeline where quick drafts/iterations run locally in the browser (FREE, ~1-3 seconds), while high-quality final renders still use RunPod SDXL.
**Problem:**
- Current image generation always hits RunPod (~$0.02/image + 10-30s cold starts)
- No instant feedback loop for creative iteration
- 100% of compute costs are cloud-based
**Solution:**
- Add WebGPU capability detection
- Integrate SD-Turbo for instant browser-based previews
- Smart routing: drafts → browser, final renders → RunPod
- Potential 70% reduction in RunPod image generation costs
**Cost Impact (projected):**
- 1,000 images/mo: $20 → $6 (save $14/mo)
- 5,000 images/mo: $100 → $30 (save $70/mo)
- 10,000 images/mo: $200 → $60 (save $140/mo)
**Browser Support:**
- Chrome/Edge: Full WebGPU (v113+)
- Firefox: Windows (July 2025)
- Safari: v26 beta
- Fallback: WASM backend for unsupported browsers
<!-- SECTION:DESCRIPTION:END -->
## Acceptance Criteria
<!-- AC:BEGIN -->
- [ ] #1 WebGPU capability detection added to clientConfig.ts
- [ ] #2 SD-Turbo model loads and runs in browser via WebGPU
- [ ] #3 ImageGenShapeUtil has Quick Preview vs High Quality toggle
- [ ] #4 Smart routing in aiOrchestrator routes drafts to browser
- [ ] #5 Fallback to WASM for browsers without WebGPU
- [ ] #6 User can generate preview images with zero cold start
- [ ] #7 RunPod only called for High Quality final renders
- [ ] #8 Model download progress indicator shown to user
- [ ] #9 Works offline after initial model download
<!-- AC:END -->
## Implementation Plan
<!-- SECTION:PLAN:BEGIN -->
## Phase 1: Foundation (Quick Wins)
### 1.1 WebGPU Capability Detection
**File:** `src/lib/clientConfig.ts`
```typescript
export async function detectWebGPUCapabilities(): Promise<{
hasWebGPU: boolean
hasF16: boolean
adapterInfo?: GPUAdapterInfo
estimatedVRAM?: number
}> {
if (!navigator.gpu) {
return { hasWebGPU: false, hasF16: false }
}
const adapter = await navigator.gpu.requestAdapter()
if (!adapter) {
return { hasWebGPU: false, hasF16: false }
}
const hasF16 = adapter.features.has('shader-f16')
const adapterInfo = await adapter.requestAdapterInfo()
return {
hasWebGPU: true,
hasF16,
adapterInfo,
estimatedVRAM: adapterInfo.memoryHeaps?.[0]?.size
}
}
```
### 1.2 Install Dependencies
```bash
npm install @anthropic-ai/sdk onnxruntime-web
# Or for transformers.js v3:
npm install @huggingface/transformers
```
### 1.3 Vite Config Updates
**File:** `vite.config.ts`
- Ensure WASM/ONNX assets are properly bundled
- Add WebGPU shader compilation support
- Configure chunk splitting for ML models
---
## Phase 2: Browser Diffusion Integration
### 2.1 Create WebGPU Diffusion Module
**New File:** `src/lib/webgpuDiffusion.ts`
```typescript
import { pipeline } from '@huggingface/transformers'
let generator: any = null
let loadingPromise: Promise<void> | null = null
export async function initSDTurbo(
onProgress?: (progress: number, status: string) => void
): Promise<void> {
if (generator) return
if (loadingPromise) return loadingPromise
loadingPromise = (async () => {
onProgress?.(0, 'Loading SD-Turbo model...')
generator = await pipeline(
'text-to-image',
'Xenova/sdxl-turbo', // or 'stabilityai/sd-turbo'
{
device: 'webgpu',
dtype: 'fp16',
progress_callback: (p) => onProgress?.(p.progress, p.status)
}
)
onProgress?.(100, 'Ready')
})()
return loadingPromise
}
export async function generateLocalImage(
prompt: string,
options?: {
width?: number
height?: number
steps?: number
seed?: number
}
): Promise<string> {
if (!generator) {
throw new Error('SD-Turbo not initialized. Call initSDTurbo() first.')
}
const result = await generator(prompt, {
width: options?.width || 512,
height: options?.height || 512,
num_inference_steps: options?.steps || 1, // SD-Turbo = 1 step
seed: options?.seed
})
// Returns base64 data URL
return result[0].image
}
export function isSDTurboReady(): boolean {
return generator !== null
}
export async function unloadSDTurbo(): Promise<void> {
generator = null
loadingPromise = null
// Force garbage collection of GPU memory
}
```
### 2.2 Create Model Download Manager
**New File:** `src/lib/modelDownloadManager.ts`
Handle progressive model downloads with:
- IndexedDB caching for persistence
- Progress tracking UI
- Resume capability for interrupted downloads
- Storage quota management
---
## Phase 3: UI Integration
### 3.1 Update ImageGenShapeUtil
**File:** `src/shapes/ImageGenShapeUtil.tsx`
Add to shape props:
```typescript
type IImageGen = TLBaseShape<"ImageGen", {
// ... existing props
generationMode: 'auto' | 'local' | 'cloud' // NEW
localModelStatus: 'not-loaded' | 'loading' | 'ready' | 'error' // NEW
localModelProgress: number // NEW (0-100)
}>
```
Add UI toggle:
```tsx
<div className="generation-mode-toggle">
<button
onClick={() => setMode('local')}
disabled={!hasWebGPU}
title={!hasWebGPU ? 'WebGPU not supported' : 'Fast preview (~1-3s)'}
>
⚡ Quick Preview
</button>
<button
onClick={() => setMode('cloud')}
title="High quality SDXL (~10-30s)"
>
✨ High Quality
</button>
</div>
```
### 3.2 Smart Generation Logic
```typescript
const generateImage = async (prompt: string) => {
const mode = shape.props.generationMode
const capabilities = await detectWebGPUCapabilities()
// Auto mode: local for iterations, cloud for final
if (mode === 'auto' || mode === 'local') {
if (capabilities.hasWebGPU && isSDTurboReady()) {
// Generate locally - instant!
const imageUrl = await generateLocalImage(prompt)
updateShape({ imageUrl, source: 'local' })
return
}
}
// Fall back to RunPod
await generateWithRunPod(prompt)
}
```
---
## Phase 4: AI Orchestrator Integration
### 4.1 Update aiOrchestrator.ts
**File:** `src/lib/aiOrchestrator.ts`
Add browser as compute target:
```typescript
type ComputeTarget = 'browser' | 'netcup' | 'runpod'
interface ImageGenerationOptions {
prompt: string
priority: 'draft' | 'final'
preferLocal?: boolean
}
async function generateImage(options: ImageGenerationOptions) {
const { hasWebGPU } = await detectWebGPUCapabilities()
// Routing logic
if (options.priority === 'draft' && hasWebGPU && isSDTurboReady()) {
return { target: 'browser', cost: 0 }
}
if (options.priority === 'final') {
return { target: 'runpod', cost: 0.02 }
}
// Fallback chain
return { target: 'runpod', cost: 0.02 }
}
```
---
## Phase 5: Advanced Features (Future)
### 5.1 Real-time img2img Refinement
- Start with browser SD-Turbo draft
- User adjusts/annotates
- Send to RunPod SDXL for final with img2img
### 5.2 Browser-based Upscaling
- Add Real-ESRGAN-lite via ONNX Runtime
- 2x/4x upscale locally before cloud render
### 5.3 Background Removal
- U2Net in browser via transformers.js
- Zero-cost background removal
### 5.4 Style Transfer
- Fast neural style transfer via WebGPU shaders
- Real-time preview on canvas
---
## Technical Considerations
### Model Sizes
| Model | Size | Load Time | Generation |
|-------|------|-----------|------------|
| SD-Turbo | ~2GB | 30-60s (first) | 1-3s |
| SD-Turbo (quantized) | ~1GB | 15-30s | 2-4s |
### Memory Management
- Unload model when tab backgrounded
- Clear GPU memory on low-memory warnings
- IndexedDB for model caching (survives refresh)
### Error Handling
- Graceful degradation to WASM if WebGPU fails
- Clear error messages for unsupported browsers
- Automatic fallback to RunPod on local failure
---
## Files to Create/Modify
**New Files:**
- `src/lib/webgpuDiffusion.ts` - SD-Turbo wrapper
- `src/lib/modelDownloadManager.ts` - Model caching
- `src/lib/webgpuCapabilities.ts` - Detection utilities
- `src/components/ModelDownloadProgress.tsx` - UI component
**Modified Files:**
- `src/lib/clientConfig.ts` - Add WebGPU detection
- `src/lib/aiOrchestrator.ts` - Add browser routing
- `src/shapes/ImageGenShapeUtil.tsx` - Add mode toggle
- `vite.config.ts` - ONNX/WASM config
- `package.json` - New dependencies
---
## Testing Checklist
- [ ] WebGPU detection works on Chrome, Edge, Firefox
- [ ] WASM fallback works on Safari/older browsers
- [ ] Model downloads and caches correctly
- [ ] Generation completes in <5s on modern GPU
- [ ] Memory cleaned up properly on unload
- [ ] Offline generation works after model cached
- [ ] RunPod fallback triggers correctly
- [ ] Cost tracking reflects local vs cloud usage
<!-- SECTION:PLAN:END -->

146
backlog/tasks/task-015 - Set-up-Cloudflare-D1-email-collector-database-for-cross-site-subscriptions.md Normal file
View File

@ -0,0 +1,146 @@
---
id: task-015
title: Set up Cloudflare D1 email-collector database for cross-site subscriptions
status: To Do
assignee: []
created_date: '2025-12-04 12:00'
updated_date: '2025-12-04 12:03'
labels:
- infrastructure
- cloudflare
- d1
- email
- cross-site
dependencies: []
priority: medium
---
## Description
<!-- SECTION:DESCRIPTION:BEGIN -->
Create a standalone Cloudflare D1 database for collecting email subscriptions across all websites (mycofi.earth, canvas.jeffemmett.com, decolonizeti.me, etc.) with easy export capabilities.
**Purpose:**
- Unified email collection from all sites
- Page-separated lists (e.g., /newsletter, /waitlist, /landing)
- Simple CSV/JSON export for email campaigns
- GDPR-compliant with unsubscribe tracking
**Sites to integrate:**
- mycofi.earth
- canvas.jeffemmett.com
- decolonizeti.me
- games.jeffemmett.com
- Future sites
**Key Features:**
- Double opt-in verification
- Source tracking (which site, which page)
- Export in multiple formats (CSV, JSON, Mailchimp)
- Basic admin dashboard or CLI for exports
- Rate limiting to prevent abuse
<!-- SECTION:DESCRIPTION:END -->
## Acceptance Criteria
<!-- AC:BEGIN -->
- [ ] #1 D1 database 'email-collector' created on Cloudflare
- [ ] #2 Schema deployed with subscribers, verification_tokens tables
- [ ] #3 POST /api/subscribe endpoint accepts email + source_site + source_page
- [ ] #4 Email verification flow with token-based double opt-in
- [ ] #5 GET /api/emails/export returns CSV with filters (site, date, verified)
- [ ] #6 Unsubscribe endpoint and tracking
- [ ] #7 Rate limiting prevents spam submissions
- [ ] #8 At least one site integrated and collecting emails
<!-- AC:END -->
## Implementation Plan
<!-- SECTION:PLAN:BEGIN -->
## Implementation Steps
### 1. Create D1 Database
```bash
wrangler d1 create email-collector
```
### 2. Create Schema File
Create `worker/email-collector-schema.sql`:
```sql
-- Email Collector Schema
-- Cross-site email subscription management
CREATE TABLE IF NOT EXISTS subscribers (
id TEXT PRIMARY KEY,
email TEXT NOT NULL,
email_hash TEXT NOT NULL, -- For duplicate checking
source_site TEXT NOT NULL,
source_page TEXT,
referrer TEXT,
ip_country TEXT,
subscribed_at TEXT DEFAULT (datetime('now')),
verified INTEGER DEFAULT 0,
verified_at TEXT,
unsubscribed INTEGER DEFAULT 0,
unsubscribed_at TEXT,
metadata TEXT -- JSON for custom fields
);
CREATE TABLE IF NOT EXISTS verification_tokens (
id TEXT PRIMARY KEY,
email TEXT NOT NULL,
token TEXT UNIQUE NOT NULL,
expires_at TEXT NOT NULL,
used INTEGER DEFAULT 0,
created_at TEXT DEFAULT (datetime('now'))
);
-- Rate limiting table
CREATE TABLE IF NOT EXISTS rate_limits (
ip_hash TEXT PRIMARY KEY,
request_count INTEGER DEFAULT 1,
window_start TEXT DEFAULT (datetime('now'))
);
-- Indexes
CREATE INDEX IF NOT EXISTS idx_subs_email_hash ON subscribers(email_hash);
CREATE INDEX IF NOT EXISTS idx_subs_site ON subscribers(source_site);
CREATE INDEX IF NOT EXISTS idx_subs_page ON subscribers(source_site, source_page);
CREATE INDEX IF NOT EXISTS idx_subs_verified ON subscribers(verified);
CREATE UNIQUE INDEX IF NOT EXISTS idx_subs_unique ON subscribers(email_hash, source_site);
CREATE INDEX IF NOT EXISTS idx_tokens_token ON verification_tokens(token);
```
### 3. Create Worker Endpoints
Create `worker/emailCollector.ts`:
```typescript
// POST /api/subscribe
// GET /api/verify/:token
// POST /api/unsubscribe
// GET /api/emails/export (auth required)
// GET /api/emails/stats
```
### 4. Export Formats
- CSV: `email,source_site,source_page,subscribed_at,verified`
- JSON: Full object array
- Mailchimp: CSV with required headers
### 5. Admin Authentication
- Use simple API key for export endpoint
- Store in Worker secret: `EMAIL_ADMIN_KEY`
### 6. Integration
Add to each site's signup form:
```javascript
fetch('https://canvas.jeffemmett.com/api/subscribe', {
method: 'POST',
body: JSON.stringify({
email: 'user@example.com',
source_site: 'mycofi.earth',
source_page: '/newsletter'
})
})
```
<!-- SECTION:PLAN:END -->

56
backlog/tasks/task-016 - Add-encryption-for-CryptID-emails-at-rest.md Normal file
View File

@ -0,0 +1,56 @@
---
id: task-016
title: Add encryption for CryptID emails at rest
status: To Do
assignee: []
created_date: '2025-12-04 12:01'
labels:
- security
- cryptid
- encryption
- privacy
- d1
dependencies:
- task-017
priority: medium
---
## Description
<!-- SECTION:DESCRIPTION:BEGIN -->
Enhance CryptID security by encrypting email addresses stored in D1 database. This protects user privacy even if the database is compromised.
**Encryption Strategy:**
- Encrypt email addresses before storing in D1
- Use Cloudflare Workers KV or environment secret for encryption key
- Store encrypted email + hash for lookups
- Decrypt only when needed (sending emails, display)
**Implementation Options:**
1. **AES-GCM encryption** with key in Worker secret
2. **Deterministic encryption** for email lookups (hash-based)
3. **Hybrid approach**: Hash for lookup index, AES for actual email
**Schema Changes:**
```sql
ALTER TABLE users ADD COLUMN email_encrypted TEXT;
ALTER TABLE users ADD COLUMN email_hash TEXT; -- For lookups
-- Migrate existing emails, then drop plaintext column
```
**Considerations:**
- Key rotation strategy
- Performance impact on lookups
- Backup/recovery implications
<!-- SECTION:DESCRIPTION:END -->
## Acceptance Criteria
<!-- AC:BEGIN -->
- [ ] #1 Encryption key securely stored in Worker secrets
- [ ] #2 Emails encrypted before D1 insert
- [ ] #3 Email lookup works via hash index
- [ ] #4 Decryption works for email display and sending
- [ ] #5 Existing emails migrated to encrypted format
- [ ] #6 Key rotation procedure documented
- [ ] #7 No plaintext emails in database
<!-- AC:END -->

63
backlog/tasks/task-017 - Deploy-CryptID-email-recovery-to-dev-branch-and-test.md Normal file
View File

@ -0,0 +1,63 @@
---
id: task-017
title: Deploy CryptID email recovery to dev branch and test
status: To Do
assignee: []
created_date: '2025-12-04 12:00'
updated_date: '2025-12-04 12:27'
labels:
- feature
- cryptid
- auth
- testing
- dev-branch
dependencies:
- task-018
- task-019
priority: high
---
## Description
<!-- SECTION:DESCRIPTION:BEGIN -->
Push the existing CryptID email recovery code changes to dev branch and test the full flow before merging to main.
**Code Changes Ready:**
- src/App.tsx - Routes for /verify-email, /link-device
- src/components/auth/CryptID.tsx - Email linking flow
- src/components/auth/Profile.tsx - Email management UI, device list
- src/css/crypto-auth.css - Styling for email/device modals
- worker/types.ts - Updated D1 types
- worker/worker.ts - Auth API routes
- worker/cryptidAuth.ts - Auth handlers (already committed)
**Test Scenarios:**
1. Link email to existing CryptID account
2. Verify email via link
3. Request device link from new device
4. Approve device link via email
5. View and revoke linked devices
6. Recover account on new device via email
<!-- SECTION:DESCRIPTION:END -->
## Acceptance Criteria
<!-- AC:BEGIN -->
- [ ] #1 All CryptID changes committed to dev branch
- [ ] #2 Worker deployed to dev environment
- [ ] #3 Link email flow works end-to-end
- [ ] #4 Email verification completes successfully
- [ ] #5 Device linking via email works
- [ ] #6 Device revocation works
- [ ] #7 Profile shows linked email and devices
- [ ] #8 No console errors in happy path
<!-- AC:END -->
## Implementation Notes
<!-- SECTION:NOTES:BEGIN -->
Branch created: `feature/cryptid-email-recovery`
Code committed and pushed to Gitea
PR available at: https://gitea.jeffemmett.com/jeffemmett/canvas-website/compare/main...feature/cryptid-email-recovery
<!-- SECTION:NOTES:END -->

111
backlog/tasks/task-018 - Create-Cloudflare-D1-cryptid-auth-database.md Normal file
View File

@ -0,0 +1,111 @@
---
id: task-018
title: Create Cloudflare D1 cryptid-auth database
status: To Do
assignee: []
created_date: '2025-12-04 12:02'
updated_date: '2025-12-04 12:27'
labels:
- infrastructure
- cloudflare
- d1
- cryptid
- auth
dependencies: []
priority: high
---
## Description
<!-- SECTION:DESCRIPTION:BEGIN -->
Create the D1 database on Cloudflare for CryptID authentication system. This is the first step before deploying the email recovery feature.
**Database Purpose:**
- Store user accounts linked to CryptID usernames
- Store device public keys for multi-device auth
- Store verification tokens for email/device linking
- Enable account recovery via verified email
**Security Considerations:**
- Emails should be encrypted at rest (task-016)
- Public keys are safe to store (not secrets)
- Tokens are time-limited and single-use
- No passwords stored (WebCrypto key-based auth)
<!-- SECTION:DESCRIPTION:END -->
## Acceptance Criteria
<!-- AC:BEGIN -->
- [ ] #1 D1 database 'cryptid-auth' created via wrangler d1 create
- [ ] #2 D1 database 'cryptid-auth-dev' created for dev environment
- [ ] #3 Database IDs added to wrangler.toml (replacing placeholders)
- [ ] #4 Schema from worker/schema.sql deployed to both databases
- [ ] #5 Verified tables exist: users, device_keys, verification_tokens
<!-- AC:END -->
## Implementation Plan
<!-- SECTION:PLAN:BEGIN -->
## Implementation Steps
### 1. Create D1 Databases
Run from local machine or Netcup (requires wrangler CLI):
```bash
cd /home/jeffe/Github/canvas-website
# Create production database
wrangler d1 create cryptid-auth
# Create dev database
wrangler d1 create cryptid-auth-dev
```
### 2. Update wrangler.toml
Replace placeholder IDs with actual database IDs from step 1:
```toml
[[d1_databases]]
binding = "CRYPTID_DB"
database_name = "cryptid-auth"
database_id = "<PROD_ID_FROM_STEP_1>"
[[env.dev.d1_databases]]
binding = "CRYPTID_DB"
database_name = "cryptid-auth-dev"
database_id = "<DEV_ID_FROM_STEP_1>"
```
### 3. Deploy Schema
```bash
# Deploy to dev first
wrangler d1 execute cryptid-auth-dev --file=./worker/schema.sql
# Then production
wrangler d1 execute cryptid-auth --file=./worker/schema.sql
```
### 4. Verify Tables
```bash
# Check dev
wrangler d1 execute cryptid-auth-dev --command="SELECT name FROM sqlite_master WHERE type='table';"
# Expected output:
# - users
# - device_keys
# - verification_tokens
```
### 5. Commit wrangler.toml Changes
```bash
git add wrangler.toml
git commit -m "chore: add D1 database IDs for cryptid-auth"
```
<!-- SECTION:PLAN:END -->
## Implementation Notes
<!-- SECTION:NOTES:BEGIN -->
Feature branch: `feature/cryptid-email-recovery`
Code is ready - waiting for D1 database creation
<!-- SECTION:NOTES:END -->

41
backlog/tasks/task-019 - Configure-CryptID-secrets-and-SendGrid-integration.md Normal file
View File

@ -0,0 +1,41 @@
---
id: task-019
title: Configure CryptID secrets and SendGrid integration
status: To Do
assignee: []
created_date: '2025-12-04 12:02'
labels:
- infrastructure
- cloudflare
- cryptid
- secrets
- sendgrid
dependencies:
- task-018
priority: high
---
## Description
<!-- SECTION:DESCRIPTION:BEGIN -->
Set up the required secrets and environment variables for CryptID email functionality on Cloudflare Workers.
**Required Secrets:**
- SENDGRID_API_KEY - For sending verification emails
- CRYPTID_EMAIL_FROM - Sender email address (e.g., auth@jeffemmett.com)
- APP_URL - Base URL for verification links (e.g., https://canvas.jeffemmett.com)
**Configuration:**
- Secrets set for both production and dev environments
- SendGrid account configured with verified sender domain
- Email templates tested
<!-- SECTION:DESCRIPTION:END -->
## Acceptance Criteria
<!-- AC:BEGIN -->
- [ ] #1 SENDGRID_API_KEY secret set via wrangler secret put
- [ ] #2 CRYPTID_EMAIL_FROM secret configured
- [ ] #3 APP_URL environment variable set in wrangler.toml
- [ ] #4 SendGrid sender domain verified (jeffemmett.com or subdomain)
- [ ] #5 Test email sends successfully from Worker
<!-- AC:END -->

63
backlog/tasks/task-024 - Open-Mapping-Collaborative-Route-Planning-Module.md Normal file
View File

@ -0,0 +1,63 @@
---
id: task-024
title: 'Open Mapping: Collaborative Route Planning Module'
status: To Do
assignee: []
created_date: '2025-12-04 14:30'
labels:
- feature
- mapping
dependencies: []
priority: high
---
## Description
<!-- SECTION:DESCRIPTION:BEGIN -->
Implement an open-source mapping and routing layer for the canvas that provides advanced route planning capabilities beyond Google Maps. Built on OpenStreetMap, OSRM/Valhalla, and MapLibre GL JS.
<!-- SECTION:DESCRIPTION:END -->
## Acceptance Criteria
<!-- AC:BEGIN -->
- [ ] #1 MapLibre GL JS integrated with tldraw canvas
- [ ] #2 OSRM routing backend deployed to Netcup
- [ ] #3 Waypoint placement and route calculation working
- [ ] #4 Multi-route comparison UI implemented
- [ ] #5 Y.js collaboration for shared route editing
- [ ] #6 Layer management panel with basemap switching
- [ ] #7 Offline tile caching via Service Worker
- [ ] #8 Budget tracking per waypoint/route
<!-- AC:END -->
## Implementation Plan
<!-- SECTION:PLAN:BEGIN -->
Phase 1 - Foundation:
- Integrate MapLibre GL JS with tldraw
- Deploy OSRM to /opt/apps/open-mapping/
- Basic waypoint and route UI
Phase 2 - Multi-Route:
- Alternative routes visualization
- Route comparison panel
- Elevation profiles
Phase 3 - Collaboration:
- Y.js integration
- Real-time cursor presence
- Share links
Phase 4 - Layers:
- Layer panel UI
- Multiple basemaps
- Custom overlays
Phase 5 - Calendar/Budget:
- Time windows on waypoints
- Cost estimation
- iCal export
Phase 6 - Optimization:
- VROOM TSP/VRP
- Offline PWA
<!-- SECTION:PLAN:END -->

23
backlog/tasks/task-high.01 - MI-Bar-UX-Modal-Fade-&-Scrollable-Try-Next.md Normal file
View File

@ -0,0 +1,23 @@
---
id: task-high.01
title: 'MI Bar UX: Modal Fade & Scrollable Try Next'
status: Done
assignee: []
created_date: '2025-12-04 06:34'
labels: []
dependencies: []
parent_task_id: task-high
---
## Description
<!-- SECTION:DESCRIPTION:BEGIN -->
Improved Mycelial Intelligence bar UX: fades when modals/popups are open, combined Tools + Follow-up suggestions into a single scrollable 'Try Next' section
<!-- SECTION:DESCRIPTION:END -->
## Acceptance Criteria
<!-- AC:BEGIN -->
- [ ] #1 MI bar fades when settings modal is open
- [ ] #2 MI bar fades when auth modal is open
- [ ] #3 Suggested tools and follow-ups in single scrollable row
<!-- AC:END -->

24
backlog/tasks/task-high.02 - CryptID-Email-Recovery-in-Settings.md Normal file
View File

@ -0,0 +1,24 @@
---
id: task-high.02
title: CryptID Email Recovery in Settings
status: Done
assignee: []
created_date: '2025-12-04 06:35'
labels: []
dependencies: []
parent_task_id: task-high
---
## Description
<!-- SECTION:DESCRIPTION:BEGIN -->
Added email linking to User Settings modal General tab - allows users to attach their email to their CryptID account for device recovery and verification
<!-- SECTION:DESCRIPTION:END -->
## Acceptance Criteria
<!-- AC:BEGIN -->
- [ ] #1 Email linking UI in General settings tab
- [ ] #2 Shows email verification status
- [ ] #3 Sends verification email on link
- [ ] #4 Dark mode aware styling
<!-- AC:END -->

38
docker-compose.yml Normal file
View File

@ -0,0 +1,38 @@
# Canvas Website Docker Compose
# Production: jeffemmett.com, www.jeffemmett.com
# Staging: staging.jeffemmett.com
services:
canvas-website:
build:
context: .
dockerfile: Dockerfile
args:
- VITE_TLDRAW_WORKER_URL=https://jeffemmett-canvas.jeffemmett.workers.dev
# Add other build args from .env if needed
container_name: canvas-website
restart: unless-stopped
labels:
- "traefik.enable=true"
- "traefik.docker.network=traefik-public"
# Single service definition (both routers use same backend)
- "traefik.http.services.canvas.loadbalancer.server.port=80"
# Production deployment (jeffemmett.com and www)
- "traefik.http.routers.canvas-prod.rule=Host(`jeffemmett.com`) || Host(`www.jeffemmett.com`)"
- "traefik.http.routers.canvas-prod.entrypoints=web"
- "traefik.http.routers.canvas-prod.service=canvas"
# Staging deployment (keep for testing)
- "traefik.http.routers.canvas-staging.rule=Host(`staging.jeffemmett.com`)"
- "traefik.http.routers.canvas-staging.entrypoints=web"
- "traefik.http.routers.canvas-staging.service=canvas"
networks:
- traefik-public
healthcheck:
test: ["CMD", "curl", "-f", "http://localhost/health"]
interval: 30s
timeout: 10s
retries: 3
networks:
traefik-public:
external: true

94
docs/WEBCRYPTO_AUTH.md
View File

@ -4,7 +4,7 @@ This document describes the complete WebCryptoAPI authentication system implemen
## Overview ## Overview
The WebCryptoAPI authentication system provides cryptographic authentication using ECDSA P-256 key pairs, challenge-response authentication, and secure key storage. It integrates with the existing ODD (Open Data Directory) framework while providing a fallback authentication mechanism. The WebCryptoAPI authentication system provides cryptographic authentication using ECDSA P-256 key pairs, challenge-response authentication, and secure key storage. This is the primary authentication mechanism for the application.
## Architecture ## Architecture
@ -23,13 +23,14 @@ The WebCryptoAPI authentication system provides cryptographic authentication usi
- User registration and login - User registration and login
- Credential verification - Credential verification
3. **Enhanced AuthService** (`src/lib/auth/authService.ts`) 3. **AuthService** (`src/lib/auth/authService.ts`)
- Integrates crypto authentication with ODD - Simplified authentication service
- Fallback mechanisms
- Session management - Session management
- Integration with CryptoAuthService
4. **UI Components** 4. **UI Components**
- `CryptoLogin.tsx` - Cryptographic authentication UI - `CryptID.tsx` - Cryptographic authentication UI
- `CryptoDebug.tsx` - Debug component for verification
- `CryptoTest.tsx` - Test component for verification - `CryptoTest.tsx` - Test component for verification
## Features ## Features
@ -41,7 +42,6 @@ The WebCryptoAPI authentication system provides cryptographic authentication usi
- **Public Key Infrastructure**: Store and verify public keys - **Public Key Infrastructure**: Store and verify public keys
- **Browser Support Detection**: Checks for WebCryptoAPI availability - **Browser Support Detection**: Checks for WebCryptoAPI availability
- **Secure Context Validation**: Ensures HTTPS requirement - **Secure Context Validation**: Ensures HTTPS requirement
- **Fallback Authentication**: Works with existing ODD system
- **Modern UI**: Responsive design with dark mode support - **Modern UI**: Responsive design with dark mode support
- **Comprehensive Testing**: Test component for verification - **Comprehensive Testing**: Test component for verification
@ -86,7 +86,7 @@ const isValid = await crypto.verifySignature(publicKey, signature, challenge);
### Feature Detection ### Feature Detection
```typescript ```typescript
const hasWebCrypto = typeof window.crypto !== 'undefined' && const hasWebCrypto = typeof window.crypto !== 'undefined' &&
typeof window.crypto.subtle !== 'undefined'; typeof window.crypto.subtle !== 'undefined';
const isSecure = window.isSecureContext; const isSecure = window.isSecureContext;
``` ```
@ -98,26 +98,26 @@ const isSecure = window.isSecureContext;
1. **Secure Context Requirement**: Only works over HTTPS 1. **Secure Context Requirement**: Only works over HTTPS
2. **ECDSA P-256**: Industry-standard elliptic curve 2. **ECDSA P-256**: Industry-standard elliptic curve
3. **Challenge-Response**: Prevents replay attacks 3. **Challenge-Response**: Prevents replay attacks
4. **Key Storage**: Public keys stored securely 4. **Key Storage**: Public keys stored securely in localStorage
5. **Input Validation**: Username format validation 5. **Input Validation**: Username format validation
6. **Error Handling**: Comprehensive error management 6. **Error Handling**: Comprehensive error management
### ⚠️ Security Notes ### ⚠️ Security Notes
1. **Private Key Storage**: Currently simplified for demo purposes 1. **Private Key Storage**: Currently uses localStorage for demo purposes
- In production, use Web Crypto API's key storage - In production, consider using Web Crypto API's non-extractable keys
- Consider hardware security modules (HSM) - Consider hardware security modules (HSM)
- Implement proper key derivation - Implement proper key derivation
2. **Session Management**: 2. **Session Management**:
- Integrates with existing ODD session system - Uses localStorage for session persistence
- Consider implementing JWT tokens - Consider implementing JWT tokens for server-side verification
- Add session expiration - Add session expiration and refresh logic
3. **Network Security**: 3. **Network Security**:
- All crypto operations happen client-side - All crypto operations happen client-side
- No private keys transmitted over network - No private keys transmitted over network
- Consider adding server-side verification - Consider adding server-side signature verification
## Usage ## Usage
@ -146,11 +146,22 @@ import { useAuth } from './context/AuthContext';
const { login, register } = useAuth(); const { login, register } = useAuth();
// The AuthService automatically tries crypto auth first, // AuthService automatically uses crypto auth
// then falls back to ODD authentication
const success = await login('username'); const success = await login('username');
``` ```
### Using the CryptID Component
```typescript
import CryptID from './components/auth/CryptID';
// Render the authentication component
<CryptID
onSuccess={() => console.log('Login successful')}
onCancel={() => console.log('Login cancelled')}
/>
```
### Testing the Implementation ### Testing the Implementation
```typescript ```typescript
@ -166,31 +177,42 @@ import CryptoTest from './components/auth/CryptoTest';
src/ src/
├── lib/ ├── lib/
│ ├── auth/ │ ├── auth/
│ │ ├── crypto.ts # WebCryptoAPI wrapper │ │ ├── crypto.ts # WebCryptoAPI wrapper
│ │ ├── cryptoAuthService.ts # High-level auth service │ │ ├── cryptoAuthService.ts # High-level auth service
│ │ ├── authService.ts # Enhanced auth service │ │ ├── authService.ts # Simplified auth service
│ │ └── account.ts # User account management │ │ ├── sessionPersistence.ts # Session storage utilities
│ │ └── types.ts # TypeScript types
│ └── utils/ │ └── utils/
│ └── browser.ts # Browser support detection │ └── browser.ts # Browser support detection
├── components/ ├── components/
│ └── auth/ │ └── auth/
│ ├── CryptoLogin.tsx # Crypto auth UI │ ├── CryptID.tsx # Main crypto auth UI
│ └── CryptoTest.tsx # Test component │ ├── CryptoDebug.tsx # Debug component
│ └── CryptoTest.tsx # Test component
├── context/
│ └── AuthContext.tsx # React context for auth state
└── css/ └── css/
└── crypto-auth.css # Styles for crypto components └── crypto-auth.css # Styles for crypto components
``` ```
## Dependencies ## Dependencies
### Required Packages ### Required Packages
- `one-webcrypto`: WebCryptoAPI polyfill (^1.0.3) - `one-webcrypto`: WebCryptoAPI polyfill (^1.0.3)
- `@oddjs/odd`: Open Data Directory framework (^0.37.2)
### Browser APIs Used ### Browser APIs Used
- `window.crypto.subtle`: WebCryptoAPI - `window.crypto.subtle`: WebCryptoAPI
- `window.localStorage`: Key storage - `window.localStorage`: Key and session storage
- `window.isSecureContext`: Security context check - `window.isSecureContext`: Security context check
## Storage
### localStorage Keys Used
- `registeredUsers`: Array of registered usernames
- `${username}_publicKey`: User's public key (Base64)
- `${username}_authData`: Authentication data (challenge, signature, timestamp)
- `session`: Current user session data
## Testing ## Testing
### Manual Testing ### Manual Testing
@ -208,6 +230,7 @@ src/
- [x] User registration - [x] User registration
- [x] User login - [x] User login
- [x] Credential verification - [x] Credential verification
- [x] Session persistence
## Troubleshooting ## Troubleshooting
@ -228,13 +251,13 @@ src/
- Try refreshing the page - Try refreshing the page
4. **"Authentication failed"** 4. **"Authentication failed"**
- Verify user exists - Verify user exists in localStorage
- Check stored credentials - Check stored credentials
- Clear browser data and retry - Clear browser data and retry
### Debug Mode ### Debug Mode
Enable debug logging by setting: Enable debug logging by opening the browser console:
```typescript ```typescript
localStorage.setItem('debug_crypto', 'true'); localStorage.setItem('debug_crypto', 'true');
``` ```
@ -242,7 +265,7 @@ localStorage.setItem('debug_crypto', 'true');
## Future Enhancements ## Future Enhancements
### Planned Improvements ### Planned Improvements
1. **Enhanced Key Storage**: Use Web Crypto API's key storage 1. **Enhanced Key Storage**: Use Web Crypto API's non-extractable keys
2. **Server-Side Verification**: Add server-side signature verification 2. **Server-Side Verification**: Add server-side signature verification
3. **Multi-Factor Authentication**: Add additional authentication factors 3. **Multi-Factor Authentication**: Add additional authentication factors
4. **Key Rotation**: Implement automatic key rotation 4. **Key Rotation**: Implement automatic key rotation
@ -254,6 +277,15 @@ localStorage.setItem('debug_crypto', 'true');
3. **Post-Quantum Cryptography**: Prepare for quantum threats 3. **Post-Quantum Cryptography**: Prepare for quantum threats
4. **Biometric Integration**: Add biometric authentication 4. **Biometric Integration**: Add biometric authentication
## Integration with Automerge Sync
The authentication system works seamlessly with the Automerge-based real-time collaboration:
- **User Identification**: Each user is identified by their username in Automerge
- **Session Management**: Sessions persist across page reloads via localStorage
- **Collaboration**: Authenticated users can join shared canvas rooms
- **Privacy**: Only authenticated users can access canvas data
## Contributing ## Contributing
When contributing to the WebCryptoAPI authentication system: When contributing to the WebCryptoAPI authentication system:
@ -269,4 +301,4 @@ When contributing to the WebCryptoAPI authentication system:
- [WebCryptoAPI Specification](https://www.w3.org/TR/WebCryptoAPI/) - [WebCryptoAPI Specification](https://www.w3.org/TR/WebCryptoAPI/)
- [ECDSA Algorithm](https://en.wikipedia.org/wiki/Elliptic_Curve_Digital_Signature_Algorithm) - [ECDSA Algorithm](https://en.wikipedia.org/wiki/Elliptic_Curve_Digital_Signature_Algorithm)
- [P-256 Curve](https://en.wikipedia.org/wiki/NIST_Curve_P-256) - [P-256 Curve](https://en.wikipedia.org/wiki/NIST_Curve_P-256)
- [Challenge-Response Authentication](https://en.wikipedia.org/wiki/Challenge%E2%80%93response_authentication) - [Challenge-Response Authentication](https://en.wikipedia.org/wiki/Challenge%E2%80%93response_authentication)

8
multmux/.dockerignore Normal file
View File

@ -0,0 +1,8 @@
node_modules
packages/*/node_modules
packages/*/dist
*.log
.git
.gitignore
README.md
infrastructure/

35
multmux/.gitignore vendored Normal file
View File

@ -0,0 +1,35 @@
# Dependencies
node_modules/
package-lock.json
# Build outputs
dist/
*.tsbuildinfo
# Logs
logs/
*.log
npm-debug.log*
yarn-debug.log*
yarn-error.log*
pm2.log
# Environment variables
.env
.env.local
.env.*.local
# IDE
.vscode/
.idea/
*.swp
*.swo
*~
# OS
.DS_Store
Thumbs.db
# PM2
ecosystem.config.js
.pm2/

32
multmux/Dockerfile Normal file
View File

@ -0,0 +1,32 @@
# mulTmux Server Dockerfile
FROM node:20-slim
# Install tmux and build dependencies for node-pty
RUN apt-get update && apt-get install -y \
tmux \
python3 \
make \
g++ \
&& rm -rf /var/lib/apt/lists/*
WORKDIR /app
# Copy workspace root files
COPY package.json ./
COPY tsconfig.json ./
# Copy packages
COPY packages/server ./packages/server
COPY packages/cli ./packages/cli
# Install dependencies (including node-pty native compilation)
RUN npm install --workspaces
# Build TypeScript
RUN npm run build
# Expose port
EXPOSE 3002
# Run the server
CMD ["node", "packages/server/dist/index.js"]

240
multmux/README.md Normal file
View File

@ -0,0 +1,240 @@
# mulTmux
A collaborative terminal tool that lets multiple users interact with the same tmux session in real-time.
## Features
- **Real-time Collaboration**: Multiple users can connect to the same terminal session
- **tmux Backend**: Leverages tmux for robust terminal multiplexing
- **Token-based Auth**: Secure invite links with expiration
- **Presence Indicators**: See who's connected to your session
- **Low Resource Usage**: ~200-300MB RAM for typical usage
- **Easy Deployment**: Works alongside existing services on your server
## Architecture
```
┌─────────────┐ ┌──────────────────┐
│ Client │ ──── WebSocket ────────> │ Server │
│ (CLI) │ (token auth) │ │
└─────────────┘ │ ┌────────────┐ │
│ │ Node.js │ │
┌─────────────┐ │ │ Backend │ │
│ Client 2 │ ──── Invite Link ──────> │ └─────┬──────┘ │
│ (CLI) │ │ │ │
└─────────────┘ │ ┌─────▼──────┐ │
│ │ tmux │ │
│ │ Sessions │ │
│ └────────────┘ │
└──────────────────┘
```
## Installation
### Server Setup
1. **Deploy to your AI server:**
```bash
cd multmux
chmod +x infrastructure/deploy.sh
./infrastructure/deploy.sh
```
This will:
- Install tmux if needed
- Build the server
- Set up PM2 for process management
- Start the server
2. **(Optional) Set up nginx reverse proxy:**
```bash
sudo cp infrastructure/nginx.conf /etc/nginx/sites-available/multmux
sudo ln -s /etc/nginx/sites-available/multmux /etc/nginx/sites-enabled/
# Edit the file to set your domain
sudo nano /etc/nginx/sites-available/multmux
sudo nginx -t
sudo systemctl reload nginx
```
### CLI Installation
**On your local machine:**
```bash
cd multmux/packages/cli
npm install
npm run build
npm link # Installs 'multmux' command globally
```
## Usage
### Create a Session
```bash
multmux create my-project --repo /path/to/repo
```
This outputs an invite link like:
```
multmux join a1b2c3d4e5f6g7h8i9j0k1l2m3n4o5p6
```
### Join a Session
```bash
multmux join a1b2c3d4e5f6g7h8i9j0k1l2m3n4o5p6
```
### List Active Sessions
```bash
multmux list
```
### Using a Remote Server
If your server is on a different machine:
```bash
# Create session
multmux create my-project --server http://your-server:3000
# Join session
multmux join <token> --server ws://your-server:3001
```
## CLI Commands
| Command | Description |
|---------|-------------|
| `multmux create <name>` | Create a new collaborative session |
| `multmux join <token>` | Join an existing session |
| `multmux list` | List all active sessions |
### Options
**create:**
- `-s, --server <url>` - Server URL (default: http://localhost:3000)
- `-r, --repo <path>` - Repository path to cd into
**join:**
- `-s, --server <url>` - WebSocket server URL (default: ws://localhost:3001)
**list:**
- `-s, --server <url>` - Server URL (default: http://localhost:3000)
## Server Management
### PM2 Commands
```bash
pm2 status # Check server status
pm2 logs multmux-server # View server logs
pm2 restart multmux-server # Restart server
pm2 stop multmux-server # Stop server
```
### Resource Usage
- **Idle**: ~100-150MB RAM
- **Per session**: ~5-10MB RAM
- **Per user**: ~1-2MB RAM
- **Typical usage**: 200-300MB RAM total
## API Reference
### HTTP API (default: port 3000)
| Endpoint | Method | Description |
|----------|--------|-------------|
| `/api/sessions` | POST | Create a new session |
| `/api/sessions` | GET | List active sessions |
| `/api/sessions/:id` | GET | Get session info |
| `/api/sessions/:id/tokens` | POST | Generate new invite token |
| `/api/health` | GET | Health check |
### WebSocket (default: port 3001)
Connect with: `ws://localhost:3001?token=<your-token>`
**Message Types:**
- `output` - Terminal output from server
- `input` - User input to terminal
- `resize` - Terminal resize event
- `presence` - User join/leave notifications
- `joined` - Connection confirmation
## Security
- **Token Expiration**: Invite tokens expire after 60 minutes (configurable)
- **Session Isolation**: Each session runs in its own tmux instance
- **Input Validation**: All terminal input is validated
- **No Persistence**: Sessions are destroyed when all users leave
## Troubleshooting
### Server won't start
Check if ports are available:
```bash
netstat -tlnp | grep -E '3000|3001'
```
### Can't connect to server
1. Check server is running: `pm2 status`
2. Check logs: `pm2 logs multmux-server`
3. Verify firewall allows ports 3000 and 3001
### Terminal not responding
1. Check WebSocket connection in browser console
2. Verify token hasn't expired
3. Restart session: `pm2 restart multmux-server`
## Development
### Project Structure
```
multmux/
├── packages/
│ ├── server/ # Backend server
│ │ ├── src/
│ │ │ ├── managers/ # Session & token management
│ │ │ ├── websocket/ # WebSocket handler
│ │ │ └── api/ # HTTP routes
│ └── cli/ # CLI client
│ ├── src/
│ │ ├── commands/ # CLI commands
│ │ ├── connection/ # WebSocket client
│ │ └── ui/ # Terminal UI
└── infrastructure/ # Deployment scripts
```
### Running in Development
**Terminal 1 - Server:**
```bash
npm run dev:server
```
**Terminal 2 - CLI:**
```bash
cd packages/cli
npm run dev -- create test-session
```
### Building
```bash
npm run build # Builds both packages
```
## License
MIT
## Contributing
Contributions welcome! Please open an issue or PR.

33
multmux/docker-compose.yml Normal file
View File

@ -0,0 +1,33 @@
version: '3.8'
services:
multmux:
build: .
container_name: multmux-server
restart: unless-stopped
environment:
- NODE_ENV=production
- PORT=3002
labels:
- "traefik.enable=true"
# HTTP router
- "traefik.http.routers.multmux.rule=Host(`terminal.jeffemmett.com`)"
- "traefik.http.routers.multmux.entrypoints=web"
- "traefik.http.services.multmux.loadbalancer.server.port=3002"
# WebSocket support - Traefik handles this automatically for HTTP/1.1 upgrades
# Enable sticky sessions for WebSocket connections
- "traefik.http.services.multmux.loadbalancer.sticky.cookie=true"
- "traefik.http.services.multmux.loadbalancer.sticky.cookie.name=multmux_session"
networks:
- traefik-public
# Health check
healthcheck:
test: ["CMD", "curl", "-f", "http://localhost:3002/api/health"]
interval: 30s
timeout: 10s
retries: 3
start_period: 10s
networks:
traefik-public:
external: true

91
multmux/infrastructure/deploy.sh Executable file
View File

@ -0,0 +1,91 @@
#!/bin/bash
# mulTmux Deployment Script for AI Server
# This script sets up mulTmux on your existing droplet
set -e
echo "🚀 mulTmux Deployment Script"
echo "============================"
echo ""
# Check if tmux is installed
if ! command -v tmux &> /dev/null; then
echo "📦 Installing tmux..."
sudo apt-get update
sudo apt-get install -y tmux
else
echo "✅ tmux is already installed"
fi
# Check if Node.js is installed
if ! command -v node &> /dev/null; then
echo "📦 Installing Node.js..."
curl -fsSL https://deb.nodesource.com/setup_20.x | sudo -E bash -
sudo apt-get install -y nodejs
else
echo "✅ Node.js is already installed ($(node --version))"
fi
# Check if npm is installed
if ! command -v npm &> /dev/null; then
echo "❌ npm is not installed. Please install npm first."
exit 1
else
echo "✅ npm is already installed ($(npm --version))"
fi
# Build the server
echo ""
echo "🔨 Building mulTmux..."
cd "$(dirname "$0")/.."
npm install
npm run build
echo ""
echo "📝 Setting up PM2 for process management..."
if ! command -v pm2 &> /dev/null; then
sudo npm install -g pm2
fi
# Create PM2 ecosystem file
cat > ecosystem.config.js << EOF
module.exports = {
apps: [{
name: 'multmux-server',
script: './packages/server/dist/index.js',
instances: 1,
autorestart: true,
watch: false,
max_memory_restart: '500M',
env: {
NODE_ENV: 'production',
PORT: 3000,
WS_PORT: 3001
}
}]
};
EOF
echo ""
echo "🚀 Starting mulTmux server with PM2..."
pm2 start ecosystem.config.js
pm2 save
pm2 startup | tail -n 1 | bash || true
echo ""
echo "✅ mulTmux deployed successfully!"
echo ""
echo "Server is running on:"
echo " HTTP API: http://localhost:3000"
echo " WebSocket: ws://localhost:3001"
echo ""
echo "Useful PM2 commands:"
echo " pm2 status - Check server status"
echo " pm2 logs multmux-server - View logs"
echo " pm2 restart multmux-server - Restart server"
echo " pm2 stop multmux-server - Stop server"
echo ""
echo "To install the CLI globally:"
echo " cd packages/cli && npm link"
echo ""

53
multmux/infrastructure/nginx.conf Normal file
View File

@ -0,0 +1,53 @@
# nginx configuration for mulTmux
# Place this in /etc/nginx/sites-available/multmux
# Then: sudo ln -s /etc/nginx/sites-available/multmux /etc/nginx/sites-enabled/
upstream multmux_api {
server localhost:3000;
}
upstream multmux_ws {
server localhost:3001;
}
server {
listen 80;
server_name your-server-domain.com; # Change this to your domain or IP
# HTTP API
location /api {
proxy_pass http://multmux_api;
proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection 'upgrade';
proxy_set_header Host $host;
proxy_cache_bypass $http_upgrade;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
}
# WebSocket
location /ws {
proxy_pass http://multmux_ws;
proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection "upgrade";
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
proxy_read_timeout 86400;
}
}
# Optional: SSL configuration (if using Let's Encrypt)
# server {
# listen 443 ssl http2;
# server_name your-server-domain.com;
#
# ssl_certificate /etc/letsencrypt/live/your-server-domain.com/fullchain.pem;
# ssl_certificate_key /etc/letsencrypt/live/your-server-domain.com/privkey.pem;
#
# # Same location blocks as above...
# }

19
multmux/package.json Normal file
View File

@ -0,0 +1,19 @@
{
"name": "multmux",
"version": "0.1.0",
"private": true,
"description": "Collaborative terminal tool with tmux backend",
"workspaces": [
"packages/*"
],
"scripts": {
"build": "npm run build -ws",
"dev:server": "npm run dev -w @multmux/server",
"dev:cli": "npm run dev -w @multmux/cli",
"start:server": "npm run start -w @multmux/server"
},
"devDependencies": {
"@types/node": "^20.0.0",
"typescript": "^5.0.0"
}
}

30
multmux/packages/cli/package.json Normal file
View File

@ -0,0 +1,30 @@
{
"name": "@multmux/cli",
"version": "0.1.0",
"description": "mulTmux CLI - collaborative terminal client",
"main": "dist/index.js",
"bin": {
"multmux": "./dist/index.js"
},
"scripts": {
"build": "tsc",
"dev": "tsx src/index.ts",
"start": "node dist/index.js"
},
"dependencies": {
"commander": "^11.1.0",
"ws": "^8.16.0",
"blessed": "^0.1.81",
"chalk": "^4.1.2",
"ora": "^5.4.1",
"node-fetch": "^2.7.0"
},
"devDependencies": {
"@types/ws": "^8.5.10",
"@types/node": "^20.0.0",
"@types/blessed": "^0.1.25",
"@types/node-fetch": "^2.6.9",
"tsx": "^4.7.0",
"typescript": "^5.0.0"
}
}

50
multmux/packages/cli/src/commands/create.ts Normal file
View File

@ -0,0 +1,50 @@
import fetch from 'node-fetch';
import chalk from 'chalk';
import ora from 'ora';
export async function createSession(
name: string,
options: { server?: string; repo?: string }
): Promise<void> {
const serverUrl = options.server || 'http://localhost:3000';
const spinner = ora('Creating session...').start();
try {
const response = await fetch(`${serverUrl}/api/sessions`, {
method: 'POST',
headers: {
'Content-Type': 'application/json',
},
body: JSON.stringify({
name,
repoPath: options.repo,
}),
});
if (!response.ok) {
throw new Error(`Failed to create session: ${response.statusText}`);
}
const data: any = await response.json();
spinner.succeed('Session created!');
console.log('');
console.log(chalk.bold('Session Details:'));
console.log(` Name: ${chalk.cyan(data.session.name)}`);
console.log(` ID: ${chalk.gray(data.session.id)}`);
console.log(` Created: ${new Date(data.session.createdAt).toLocaleString()}`);
console.log('');
console.log(chalk.bold('To join this session:'));
console.log(chalk.green(` ${data.inviteUrl}`));
console.log('');
console.log(chalk.bold('Or share this token:'));
console.log(` ${chalk.yellow(data.token)}`);
console.log('');
console.log(chalk.dim('Token expires in 60 minutes'));
} catch (error) {
spinner.fail('Failed to create session');
console.error(chalk.red((error as Error).message));
process.exit(1);
}
}

45
multmux/packages/cli/src/commands/join.ts Normal file
View File

@ -0,0 +1,45 @@
import chalk from 'chalk';
import ora from 'ora';
import { WebSocketClient } from '../connection/WebSocketClient';
import { TerminalUI } from '../ui/Terminal';
export async function joinSession(
token: string,
options: { server?: string }
): Promise<void> {
const serverUrl = options.server || 'ws://localhost:3001';
const spinner = ora('Connecting to session...').start();
try {
const client = new WebSocketClient(serverUrl, token);
// Wait for connection
await client.connect();
spinner.succeed('Connected!');
// Wait a moment for the 'joined' event
await new Promise((resolve) => {
client.once('joined', resolve);
setTimeout(resolve, 1000); // Fallback timeout
});
console.log(chalk.green('\nJoined session! Press ESC or Ctrl-C to exit.\n'));
// Create terminal UI
const ui = new TerminalUI(client);
// Handle errors
client.on('error', (error: Error) => {
console.error(chalk.red('\nConnection error:'), error.message);
});
client.on('reconnect-failed', () => {
console.error(chalk.red('\nFailed to reconnect. Exiting...'));
process.exit(1);
});
} catch (error) {
spinner.fail('Failed to connect');
console.error(chalk.red((error as Error).message));
process.exit(1);
}
}

38
multmux/packages/cli/src/commands/list.ts Normal file
View File

@ -0,0 +1,38 @@
import fetch from 'node-fetch';
import chalk from 'chalk';
import ora from 'ora';
export async function listSessions(options: { server?: string }): Promise<void> {
const serverUrl = options.server || 'http://localhost:3000';
const spinner = ora('Fetching sessions...').start();
try {
const response = await fetch(`${serverUrl}/api/sessions`);
if (!response.ok) {
throw new Error(`Failed to fetch sessions: ${response.statusText}`);
}
const data: any = await response.json();
spinner.stop();
if (data.sessions.length === 0) {
console.log(chalk.yellow('No active sessions found.'));
return;
}
console.log(chalk.bold(`\nActive Sessions (${data.sessions.length}):\n`));
data.sessions.forEach((session: any) => {
console.log(chalk.cyan(` ${session.name}`));
console.log(` ID: ${chalk.gray(session.id)}`);
console.log(` Clients: ${session.activeClients}`);
console.log(` Created: ${new Date(session.createdAt).toLocaleString()}`);
console.log('');
});
} catch (error) {
spinner.fail('Failed to fetch sessions');
console.error(chalk.red((error as Error).message));
process.exit(1);
}
}

120
multmux/packages/cli/src/connection/WebSocketClient.ts Normal file
View File

@ -0,0 +1,120 @@
import WebSocket from 'ws';
import { EventEmitter } from 'events';
export interface TerminalMessage {
type: 'output' | 'input' | 'resize' | 'join' | 'leave' | 'presence' | 'joined' | 'error';
data?: any;
clientId?: string;
timestamp?: number;
sessionId?: string;
sessionName?: string;
message?: string;
}
export class WebSocketClient extends EventEmitter {
private ws: WebSocket | null = null;
private reconnectAttempts = 0;
private maxReconnectAttempts = 5;
constructor(private url: string, private token: string) {
super();
}
connect(): Promise<void> {
return new Promise((resolve, reject) => {
const wsUrl = `${this.url}?token=${this.token}`;
this.ws = new WebSocket(wsUrl);
this.ws.on('open', () => {
this.reconnectAttempts = 0;
this.emit('connected');
resolve();
});
this.ws.on('message', (data) => {
try {
const message: TerminalMessage = JSON.parse(data.toString());
this.handleMessage(message);
} catch (error) {
console.error('Failed to parse message:', error);
}
});
this.ws.on('close', () => {
this.emit('disconnected');
this.attemptReconnect();
});
this.ws.on('error', (error) => {
this.emit('error', error);
reject(error);
});
});
}
private handleMessage(message: TerminalMessage): void {
switch (message.type) {
case 'output':
this.emit('output', message.data);
break;
case 'joined':
this.emit('joined', {
sessionId: message.sessionId,
sessionName: message.sessionName,
clientId: message.clientId,
});
break;
case 'presence':
this.emit('presence', message.data);
break;
case 'error':
this.emit('error', new Error(message.message || 'Unknown error'));
break;
}
}
sendInput(data: string): void {
if (this.ws && this.ws.readyState === WebSocket.OPEN) {
this.ws.send(
JSON.stringify({
type: 'input',
data,
timestamp: Date.now(),
})
);
}
}
resize(cols: number, rows: number): void {
if (this.ws && this.ws.readyState === WebSocket.OPEN) {
this.ws.send(
JSON.stringify({
type: 'resize',
data: { cols, rows },
timestamp: Date.now(),
})
);
}
}
disconnect(): void {
if (this.ws) {
this.ws.close();
this.ws = null;
}
}
private attemptReconnect(): void {
if (this.reconnectAttempts < this.maxReconnectAttempts) {
this.reconnectAttempts++;
setTimeout(() => {
this.emit('reconnecting', this.reconnectAttempts);
this.connect().catch(() => {
// Reconnection failed, will retry
});
}, 1000 * this.reconnectAttempts);
} else {
this.emit('reconnect-failed');
}
}
}

34
multmux/packages/cli/src/index.ts Normal file
View File

@ -0,0 +1,34 @@
#!/usr/bin/env node
import { Command } from 'commander';
import { createSession } from './commands/create';
import { joinSession } from './commands/join';
import { listSessions } from './commands/list';
const program = new Command();
program
.name('multmux')
.description('Collaborative terminal tool with tmux backend')
.version('0.1.0');
program
.command('create <name>')
.description('Create a new collaborative session')
.option('-s, --server <url>', 'Server URL', 'http://localhost:3000')
.option('-r, --repo <path>', 'Repository path to use')
.action(createSession);
program
.command('join <token>')
.description('Join an existing session with a token')
.option('-s, --server <url>', 'WebSocket server URL', 'ws://localhost:3001')
.action(joinSession);
program
.command('list')
.description('List active sessions')
.option('-s, --server <url>', 'Server URL', 'http://localhost:3000')
.action(listSessions);
program.parse();

154
multmux/packages/cli/src/ui/Terminal.ts Normal file
View File

@ -0,0 +1,154 @@
import blessed from 'blessed';
import { WebSocketClient } from '../connection/WebSocketClient';
export class TerminalUI {
private screen: blessed.Widgets.Screen;
private terminal: blessed.Widgets.BoxElement;
private statusBar: blessed.Widgets.BoxElement;
private buffer: string = '';
constructor(private client: WebSocketClient) {
// Create screen
this.screen = blessed.screen({
smartCSR: true,
title: 'mulTmux',
});
// Status bar
this.statusBar = blessed.box({
top: 0,
left: 0,
width: '100%',
height: 1,
style: {
fg: 'white',
bg: 'blue',
},
content: ' mulTmux - Connecting...',
});
// Terminal output
this.terminal = blessed.box({
top: 1,
left: 0,
width: '100%',
height: '100%-1',
scrollable: true,
alwaysScroll: true,
scrollbar: {
style: {
bg: 'blue',
},
},
keys: true,
vi: true,
mouse: true,
content: '',
});
this.screen.append(this.statusBar);
this.screen.append(this.terminal);
// Focus terminal
this.terminal.focus();
// Setup event handlers
this.setupEventHandlers();
// Render
this.screen.render();
}
private setupEventHandlers(): void {
// Handle terminal output from server
this.client.on('output', (data: string) => {
this.buffer += data;
this.terminal.setContent(this.buffer);
this.terminal.setScrollPerc(100);
this.screen.render();
});
// Handle connection events
this.client.on('connected', () => {
this.updateStatus('Connected', 'green');
});
this.client.on('joined', (info: any) => {
this.updateStatus(`Session: ${info.sessionName} (${info.clientId.slice(0, 8)})`, 'green');
});
this.client.on('disconnected', () => {
this.updateStatus('Disconnected', 'red');
});
this.client.on('reconnecting', (attempt: number) => {
this.updateStatus(`Reconnecting (${attempt}/5)...`, 'yellow');
});
this.client.on('presence', (data: any) => {
if (data.action === 'join') {
this.showNotification(`User joined (${data.totalClients} online)`);
} else if (data.action === 'leave') {
this.showNotification(`User left (${data.totalClients} online)`);
}
});
// Handle keyboard input
this.screen.on('keypress', (ch: string, key: any) => {
if (key.name === 'escape' || (key.ctrl && key.name === 'c')) {
this.close();
return;
}
// Send input to server
if (ch) {
this.client.sendInput(ch);
} else if (key.name) {
// Handle special keys
const specialKeys: { [key: string]: string } = {
enter: '\r',
backspace: '\x7f',
tab: '\t',
up: '\x1b[A',
down: '\x1b[B',
right: '\x1b[C',
left: '\x1b[D',
};
if (specialKeys[key.name]) {
this.client.sendInput(specialKeys[key.name]);
}
}
});
// Handle resize
this.screen.on('resize', () => {
const { width, height } = this.terminal;
this.client.resize(width as number, (height as number) - 1);
});
// Quit on Ctrl-C
this.screen.key(['C-c'], () => {
this.close();
});
}
private updateStatus(text: string, color: string = 'blue'): void {
this.statusBar.style.bg = color;
this.statusBar.setContent(` mulTmux - ${text}`);
this.screen.render();
}
private showNotification(text: string): void {
// Append notification to buffer
this.buffer += `\n[mulTmux] ${text}\n`;
this.terminal.setContent(this.buffer);
this.screen.render();
}
close(): void {
this.client.disconnect();
this.screen.destroy();
process.exit(0);
}
}

8
multmux/packages/cli/tsconfig.json Normal file
View File

@ -0,0 +1,8 @@
{
"extends": "../../tsconfig.json",
"compilerOptions": {
"outDir": "./dist",
"rootDir": "./src"
},
"include": ["src/**/*"]
}

26
multmux/packages/server/package.json Normal file
View File

@ -0,0 +1,26 @@
{
"name": "@multmux/server",
"version": "0.1.0",
"description": "mulTmux server - collaborative terminal backend",
"main": "dist/index.js",
"scripts": {
"build": "tsc",
"dev": "tsx watch src/index.ts",
"start": "node dist/index.js"
},
"dependencies": {
"express": "^4.18.0",
"ws": "^8.16.0",
"node-pty": "^1.0.0",
"nanoid": "^3.3.7",
"cors": "^2.8.5"
},
"devDependencies": {
"@types/express": "^4.17.21",
"@types/ws": "^8.5.10",
"@types/node": "^20.0.0",
"@types/cors": "^2.8.17",
"tsx": "^4.7.0",
"typescript": "^5.0.0"
}
}

116
multmux/packages/server/src/api/routes.ts Normal file
View File

@ -0,0 +1,116 @@
import { Router } from 'express';
import { SessionManager } from '../managers/SessionManager';
import { TokenManager } from '../managers/TokenManager';
export function createRouter(
sessionManager: SessionManager,
tokenManager: TokenManager
): Router {
const router = Router();
// Create a new session
router.post('/sessions', async (req, res) => {
try {
const { name, repoPath } = req.body;
if (!name || typeof name !== 'string') {
return res.status(400).json({ error: 'Session name is required' });
}
const session = await sessionManager.createSession(name, repoPath);
const token = tokenManager.generateToken(session.id, 60, 'write');
res.json({
session: {
id: session.id,
name: session.name,
createdAt: session.createdAt,
},
token,
inviteUrl: `multmux join ${token}`,
});
} catch (error) {
console.error('Failed to create session:', error);
res.status(500).json({ error: 'Failed to create session' });
}
});
// List active sessions
router.get('/sessions', (req, res) => {
const sessions = sessionManager.listSessions();
res.json({
sessions: sessions.map((s) => ({
id: s.id,
name: s.name,
createdAt: s.createdAt,
activeClients: s.clients.size,
})),
});
});
// Get session info
router.get('/sessions/:id', (req, res) => {
const session = sessionManager.getSession(req.params.id);
if (!session) {
return res.status(404).json({ error: 'Session not found' });
}
res.json({
id: session.id,
name: session.name,
createdAt: session.createdAt,
activeClients: session.clients.size,
});
});
// Join an existing session (generates a new token and returns session info)
router.post('/sessions/:id/join', (req, res) => {
const session = sessionManager.getSession(req.params.id);
if (!session) {
return res.status(404).json({ error: 'Session not found' });
}
// Generate a new token for this joining client
const token = tokenManager.generateToken(session.id, 60, 'write');
res.json({
id: session.id,
name: session.name,
token,
createdAt: session.createdAt,
activeClients: session.clients.size,
});
});
// Generate new invite token for existing session
router.post('/sessions/:id/tokens', (req, res) => {
const session = sessionManager.getSession(req.params.id);
if (!session) {
return res.status(404).json({ error: 'Session not found' });
}
const { expiresInMinutes = 60, permissions = 'write' } = req.body;
const token = tokenManager.generateToken(session.id, expiresInMinutes, permissions);
res.json({
token,
inviteUrl: `multmux join ${token}`,
expiresInMinutes,
permissions,
});
});
// Health check
router.get('/health', (req, res) => {
res.json({
status: 'ok',
activeSessions: sessionManager.listSessions().length,
activeTokens: tokenManager.getActiveTokens(),
});
});
return router;
}

55
multmux/packages/server/src/index.ts Normal file
View File

@ -0,0 +1,55 @@
import express from 'express';
import { createServer } from 'http';
import { WebSocketServer } from 'ws';
import cors from 'cors';
import { SessionManager } from './managers/SessionManager';
import { TokenManager } from './managers/TokenManager';
import { TerminalHandler } from './websocket/TerminalHandler';
import { createRouter } from './api/routes';
const PORT = process.env.PORT || 3002;
async function main() {
// Initialize managers
const sessionManager = new SessionManager();
const tokenManager = new TokenManager();
const terminalHandler = new TerminalHandler(sessionManager, tokenManager);
// HTTP API Server
const app = express();
app.use(cors());
app.use(express.json());
app.use('/api', createRouter(sessionManager, tokenManager));
// Create HTTP server to share with WebSocket
const server = createServer(app);
// WebSocket Server on same port, handles upgrade requests
const wss = new WebSocketServer({ server, path: '/ws' });
wss.on('connection', (ws, req) => {
// Extract token from query string
const url = new URL(req.url || '', `http://localhost:${PORT}`);
const token = url.searchParams.get('token');
if (!token) {
ws.send(JSON.stringify({ type: 'error', message: 'Token required' }));
ws.close();
return;
}
terminalHandler.handleConnection(ws, token);
});
server.listen(PORT, () => {
console.log('');
console.log('mulTmux server is ready!');
console.log(`API: http://localhost:${PORT}/api`);
console.log(`WebSocket: ws://localhost:${PORT}/ws`);
});
}
main().catch((error) => {
console.error('Failed to start server:', error);
process.exit(1);
});

114
multmux/packages/server/src/managers/SessionManager.ts Normal file
View File

@ -0,0 +1,114 @@
import { spawn, ChildProcess } from 'child_process';
import * as pty from 'node-pty';
import { Session } from '../types';
import { nanoid } from 'nanoid';
export class SessionManager {
private sessions: Map<string, Session> = new Map();
private terminals: Map<string, pty.IPty> = new Map();
async createSession(name: string, repoPath?: string): Promise<Session> {
const id = nanoid(16);
const tmuxSessionName = `multmux-${id}`;
const session: Session = {
id,
name,
createdAt: new Date(),
tmuxSessionName,
clients: new Set(),
repoPath,
};
this.sessions.set(id, session);
// Create tmux session
await this.createTmuxSession(tmuxSessionName, repoPath);
// Attach to tmux session with pty
const terminal = pty.spawn('tmux', ['attach-session', '-t', tmuxSessionName], {
name: 'xterm-256color',
cols: 80,
rows: 24,
cwd: repoPath || process.cwd(),
env: process.env as { [key: string]: string },
});
this.terminals.set(id, terminal);
return session;
}
private async createTmuxSession(name: string, cwd?: string): Promise<void> {
return new Promise((resolve, reject) => {
const args = ['new-session', '-d', '-s', name];
if (cwd) {
args.push('-c', cwd);
}
const proc = spawn('tmux', args);
proc.on('exit', (code) => {
if (code === 0) {
resolve();
} else {
reject(new Error(`Failed to create tmux session: exit code ${code}`));
}
});
});
}
getSession(id: string): Session | undefined {
return this.sessions.get(id);
}
getTerminal(sessionId: string): pty.IPty | undefined {
return this.terminals.get(sessionId);
}
addClient(sessionId: string, clientId: string): void {
const session = this.sessions.get(sessionId);
if (session) {
session.clients.add(clientId);
}
}
removeClient(sessionId: string, clientId: string): void {
const session = this.sessions.get(sessionId);
if (session) {
session.clients.delete(clientId);
// Clean up session if no clients left
if (session.clients.size === 0) {
this.destroySession(sessionId);
}
}
}
private async destroySession(sessionId: string): Promise<void> {
const session = this.sessions.get(sessionId);
if (!session) return;
const terminal = this.terminals.get(sessionId);
if (terminal) {
terminal.kill();
this.terminals.delete(sessionId);
}
// Kill tmux session
spawn('tmux', ['kill-session', '-t', session.tmuxSessionName]);
this.sessions.delete(sessionId);
}
listSessions(): Session[] {
return Array.from(this.sessions.values());
}
resizeTerminal(sessionId: string, cols: number, rows: number): void {
const terminal = this.terminals.get(sessionId);
if (terminal) {
terminal.resize(cols, rows);
}
}
}

50
multmux/packages/server/src/managers/TokenManager.ts Normal file
View File

@ -0,0 +1,50 @@
import { nanoid } from 'nanoid';
import { SessionToken } from '../types';
export class TokenManager {
private tokens: Map<string, SessionToken> = new Map();
generateToken(
sessionId: string,
expiresInMinutes: number = 60,
permissions: 'read' | 'write' = 'write'
): string {
const token = nanoid(32);
const expiresAt = new Date(Date.now() + expiresInMinutes * 60 * 1000);
this.tokens.set(token, {
token,
sessionId,
expiresAt,
permissions,
});
// Clean up expired token after expiration
setTimeout(() => this.tokens.delete(token), expiresInMinutes * 60 * 1000);
return token;
}
validateToken(token: string): SessionToken | null {
const sessionToken = this.tokens.get(token);
if (!sessionToken) {
return null;
}
if (sessionToken.expiresAt < new Date()) {
this.tokens.delete(token);
return null;
}
return sessionToken;
}
revokeToken(token: string): void {
this.tokens.delete(token);
}
getActiveTokens(): number {
return this.tokens.size;
}
}

29
multmux/packages/server/src/types/index.ts Normal file
View File

@ -0,0 +1,29 @@
export interface Session {
id: string;
name: string;
createdAt: Date;
tmuxSessionName: string;
clients: Set<string>;
repoPath?: string;
}
export interface SessionToken {
token: string;
sessionId: string;
expiresAt: Date;
permissions: 'read' | 'write';
}
export interface ClientConnection {
id: string;
sessionId: string;
username?: string;
permissions: 'read' | 'write';
}
export interface TerminalMessage {
type: 'output' | 'input' | 'resize' | 'join' | 'leave' | 'presence';
data: any;
clientId?: string;
timestamp: number;
}

175
multmux/packages/server/src/websocket/TerminalHandler.ts Normal file
View File

@ -0,0 +1,175 @@
import { WebSocket } from 'ws';
import { nanoid } from 'nanoid';
import { SessionManager } from '../managers/SessionManager';
import { TokenManager } from '../managers/TokenManager';
import { TerminalMessage, ClientConnection } from '../types';
export class TerminalHandler {
private clients: Map<string, { ws: WebSocket; connection: ClientConnection }> = new Map();
constructor(
private sessionManager: SessionManager,
private tokenManager: TokenManager
) {}
handleConnection(ws: WebSocket, token: string): void {
// Validate token
const sessionToken = this.tokenManager.validateToken(token);
if (!sessionToken) {
ws.send(JSON.stringify({ type: 'error', message: 'Invalid or expired token' }));
ws.close();
return;
}
// Verify session exists
const session = this.sessionManager.getSession(sessionToken.sessionId);
if (!session) {
ws.send(JSON.stringify({ type: 'error', message: 'Session not found' }));
ws.close();
return;
}
const clientId = nanoid(16);
const connection: ClientConnection = {
id: clientId,
sessionId: sessionToken.sessionId,
permissions: sessionToken.permissions,
};
this.clients.set(clientId, { ws, connection });
this.sessionManager.addClient(sessionToken.sessionId, clientId);
// Attach terminal output to WebSocket
const terminal = this.sessionManager.getTerminal(sessionToken.sessionId);
if (terminal) {
const onData = (data: string) => {
const message: TerminalMessage = {
type: 'output',
data,
timestamp: Date.now(),
};
ws.send(JSON.stringify(message));
};
const dataListener = terminal.onData(onData);
// Clean up on disconnect
ws.on('close', () => {
dataListener.dispose();
this.handleDisconnect(clientId);
});
}
// Send join confirmation
ws.send(
JSON.stringify({
type: 'joined',
sessionId: session.id,
sessionName: session.name,
clientId,
})
);
// Broadcast presence
this.broadcastToSession(sessionToken.sessionId, {
type: 'presence',
data: {
action: 'join',
clientId,
totalClients: session.clients.size,
},
timestamp: Date.now(),
});
// Handle incoming messages
ws.on('message', (data) => {
this.handleMessage(clientId, data.toString());
});
}
private handleMessage(clientId: string, rawMessage: string): void {
const client = this.clients.get(clientId);
if (!client) return;
try {
const message: TerminalMessage = JSON.parse(rawMessage);
switch (message.type) {
case 'input':
this.handleInput(client.connection, message.data);
break;
case 'resize':
this.handleResize(client.connection, message.data);
break;
}
} catch (error) {
console.error('Failed to parse message:', error);
}
}
private handleInput(connection: ClientConnection, data: string): void {
if (connection.permissions !== 'write') {
return; // Read-only clients can't send input
}
const terminal = this.sessionManager.getTerminal(connection.sessionId);
if (terminal) {
terminal.write(data);
}
// Broadcast input to other clients for cursor tracking
this.broadcastToSession(
connection.sessionId,
{
type: 'input',
data,
clientId: connection.id,
timestamp: Date.now(),
},
connection.id // Exclude sender
);
}
private handleResize(connection: ClientConnection, data: { cols: number; rows: number }): void {
this.sessionManager.resizeTerminal(connection.sessionId, data.cols, data.rows);
}
private handleDisconnect(clientId: string): void {
const client = this.clients.get(clientId);
if (!client) return;
this.sessionManager.removeClient(client.connection.sessionId, clientId);
this.clients.delete(clientId);
// Broadcast leave
const session = this.sessionManager.getSession(client.connection.sessionId);
if (session) {
this.broadcastToSession(client.connection.sessionId, {
type: 'presence',
data: {
action: 'leave',
clientId,
totalClients: session.clients.size,
},
timestamp: Date.now(),
});
}
}
private broadcastToSession(
sessionId: string,
message: TerminalMessage,
excludeClientId?: string
): void {
const session = this.sessionManager.getSession(sessionId);
if (!session) return;
const messageStr = JSON.stringify(message);
for (const [clientId, client] of this.clients.entries()) {
if (client.connection.sessionId === sessionId && clientId !== excludeClientId) {
client.ws.send(messageStr);
}
}
}
}

8
multmux/packages/server/tsconfig.json Normal file
View File

@ -0,0 +1,8 @@
{
"extends": "../../tsconfig.json",
"compilerOptions": {
"outDir": "./dist",
"rootDir": "./src"
},
"include": ["src/**/*"]
}

18
multmux/tsconfig.json Normal file
View File

@ -0,0 +1,18 @@
{
"compilerOptions": {
"target": "ES2022",
"module": "commonjs",
"lib": ["ES2022"],
"outDir": "./dist",
"rootDir": "./src",
"strict": true,
"esModuleInterop": true,
"skipLibCheck": true,
"forceConsistentCasingInFileNames": true,
"resolveJsonModule": true,
"declaration": true,
"declarationMap": true,
"sourceMap": true
},
"exclude": ["node_modules", "dist"]
}

37
nginx.conf Normal file
View File

@ -0,0 +1,37 @@
server {
listen 80;
server_name _;
root /usr/share/nginx/html;
index index.html;
# Gzip compression
gzip on;
gzip_vary on;
gzip_min_length 1024;
gzip_proxied expired no-cache no-store private auth;
gzip_types text/plain text/css text/xml text/javascript application/x-javascript application/xml application/javascript application/json;
gzip_disable "MSIE [1-6]\.";
# Security headers
add_header X-Frame-Options "SAMEORIGIN" always;
add_header X-Content-Type-Options "nosniff" always;
add_header X-XSS-Protection "1; mode=block" always;
add_header Referrer-Policy "strict-origin-when-cross-origin" always;
# Cache static assets
location ~* \.(js|css|png|jpg|jpeg|gif|ico|svg|woff|woff2|ttf|eot)$ {
expires 1y;
add_header Cache-Control "public, immutable";
}
# Handle SPA routing - all routes serve index.html
location / {
try_files $uri $uri/ /index.html;
}
# Health check endpoint
location /health {
return 200 'OK';
add_header Content-Type text/plain;
}
}

7608
package-lock.json generated
View File

File diff suppressed because it is too large Load Diff

26
package.json
View File

@ -3,8 +3,11 @@
"version": "1.0.0", "version": "1.0.0",
"description": "Jeff Emmett's personal website", "description": "Jeff Emmett's personal website",
"type": "module", "type": "module",
"workspaces": [
"multmux/packages/*"
],
"scripts": { "scripts": {
"dev": "concurrently --kill-others --names client,worker --prefix-colors blue,red \"npm run dev:client\" \"npm run dev:worker:local\"", "dev": "concurrently --kill-others --names client,worker,multmux --prefix-colors blue,red,magenta \"npm run dev:client\" \"npm run dev:worker:local\" \"npm run multmux:dev:server\"",
"dev:client": "vite --host 0.0.0.0 --port 5173", "dev:client": "vite --host 0.0.0.0 --port 5173",
"dev:worker": "wrangler dev --config wrangler.dev.toml --remote --port 5172", "dev:worker": "wrangler dev --config wrangler.dev.toml --remote --port 5172",
"dev:worker:local": "wrangler dev --config wrangler.dev.toml --port 5172 --ip 0.0.0.0", "dev:worker:local": "wrangler dev --config wrangler.dev.toml --port 5172 --ip 0.0.0.0",
@ -15,7 +18,12 @@
"deploy:pages": "tsc && vite build", "deploy:pages": "tsc && vite build",
"deploy:worker": "wrangler deploy", "deploy:worker": "wrangler deploy",
"deploy:worker:dev": "wrangler deploy --config wrangler.dev.toml", "deploy:worker:dev": "wrangler deploy --config wrangler.dev.toml",
"types": "tsc --noEmit" "types": "tsc --noEmit",
"multmux:install": "npm install --workspaces",
"multmux:build": "npm run build --workspace=@multmux/server --workspace=@multmux/cli",
"multmux:dev:server": "npm run dev --workspace=@multmux/server",
"multmux:dev:cli": "npm run dev --workspace=@multmux/cli",
"multmux:start": "npm run start --workspace=@multmux/server"
}, },
"keywords": [], "keywords": [],
"author": "Jeff Emmett", "author": "Jeff Emmett",
@ -29,15 +37,16 @@
"@chengsokdara/use-whisper": "^0.2.0", "@chengsokdara/use-whisper": "^0.2.0",
"@daily-co/daily-js": "^0.60.0", "@daily-co/daily-js": "^0.60.0",
"@daily-co/daily-react": "^0.20.0", "@daily-co/daily-react": "^0.20.0",
"@oddjs/odd": "^0.37.2", "@mdxeditor/editor": "^3.51.0",
"@tldraw/assets": "^3.15.4", "@tldraw/assets": "^3.15.4",
"@tldraw/tldraw": "^3.15.4", "@tldraw/tldraw": "^3.15.4",
"@tldraw/tlschema": "^3.15.4", "@tldraw/tlschema": "^3.15.4",
"@types/markdown-it": "^14.1.1", "@types/markdown-it": "^14.1.1",
"@types/marked": "^5.0.2", "@types/marked": "^5.0.2",
"@uiw/react-md-editor": "^4.0.5", "@uiw/react-md-editor": "^4.0.5",
"@vercel/analytics": "^1.2.2",
"@xenova/transformers": "^2.17.2", "@xenova/transformers": "^2.17.2",
"@xterm/addon-fit": "^0.10.0",
"@xterm/xterm": "^5.5.0",
"ai": "^4.1.0", "ai": "^4.1.0",
"ajv": "^8.17.1", "ajv": "^8.17.1",
"cherry-markdown": "^0.8.57", "cherry-markdown": "^0.8.57",
@ -62,9 +71,9 @@
"react-markdown": "^10.1.0", "react-markdown": "^10.1.0",
"react-router-dom": "^7.0.2", "react-router-dom": "^7.0.2",
"recoil": "^0.7.7", "recoil": "^0.7.7",
"sharp": "^0.33.5",
"tldraw": "^3.15.4", "tldraw": "^3.15.4",
"use-whisper": "^0.0.1", "use-whisper": "^0.0.1",
"vercel": "^39.1.1",
"webcola": "^3.4.0", "webcola": "^3.4.0",
"webnative": "^0.36.3" "webnative": "^0.36.3"
}, },
@ -84,6 +93,11 @@
"wrangler": "^4.33.2" "wrangler": "^4.33.2"
}, },
"engines": { "engines": {
"node": ">=18.0.0" "node": ">=20.0.0"
},
"overrides": {
"@xenova/transformers": {
"sharp": "0.33.5"
}
} }
} }

63
runpod-configs/extra_model_paths.yaml Normal file
View File

@ -0,0 +1,63 @@
# ComfyUI Model Paths Configuration
# Updated to include /runpod-volume/ paths for all model types
# This allows models to be loaded from the network volume for faster cold starts
comfyui:
base_path: /ComfyUI/
is_default: true
# Checkpoints - check network volume first, then local
checkpoints: |
/runpod-volume/models/checkpoints/
models/checkpoints/
# CLIP models
clip: |
/runpod-volume/models/clip/
models/clip/
# CLIP Vision models (e.g., clip_vision_h.safetensors)
clip_vision: |
/runpod-volume/models/clip_vision/
models/clip_vision/
# Config files
configs: models/configs/
# ControlNet models
controlnet: |
/runpod-volume/models/controlnet/
models/controlnet/
# Diffusion models (Wan2.2 model files)
diffusion_models: |
/runpod-volume/models/diffusion_models/
/runpod-volume/models/
models/diffusion_models/
models/unet/
# Text embeddings
embeddings: |
/runpod-volume/models/embeddings/
models/embeddings/
# LoRA models
loras: |
/runpod-volume/loras/
/runpod-volume/models/loras/
models/loras/
# Text encoders (e.g., umt5-xxl-enc-bf16.safetensors)
text_encoders: |
/runpod-volume/models/text_encoders/
models/text_encoders/
# Upscale models
upscale_models: |
/runpod-volume/models/upscale_models/
models/upscale_models/
# VAE models (e.g., Wan2_1_VAE_bf16.safetensors)
vae: |
/runpod-volume/models/vae/
models/vae/

143
runpod-configs/setup_network_volume.sh Normal file
View File

@ -0,0 +1,143 @@
#!/bin/bash
# Script to set up the RunPod network volume with Wan2.2 models
# Run this once on a GPU pod with the network volume attached
echo "=== Setting up RunPod Network Volume for Wan2.2 ==="
# Create directory structure
echo "Creating directory structure..."
mkdir -p /runpod-volume/models/diffusion_models
mkdir -p /runpod-volume/models/vae
mkdir -p /runpod-volume/models/text_encoders
mkdir -p /runpod-volume/models/clip_vision
mkdir -p /runpod-volume/loras
# Check current disk usage
echo "Current network volume usage:"
df -h /runpod-volume
# List what's already on the volume
echo ""
echo "Current contents of /runpod-volume:"
ls -la /runpod-volume/
echo ""
echo "Current contents of /runpod-volume/models/ (if exists):"
ls -la /runpod-volume/models/ 2>/dev/null || echo "(empty or doesn't exist)"
# Check if models exist in the Docker image
echo ""
echo "Models in Docker image /ComfyUI/models/diffusion_models/:"
ls -la /ComfyUI/models/diffusion_models/ 2>/dev/null || echo "(not found)"
echo ""
echo "Models in Docker image /ComfyUI/models/vae/:"
ls -la /ComfyUI/models/vae/ 2>/dev/null || echo "(not found)"
echo ""
echo "Models in Docker image /ComfyUI/models/text_encoders/:"
ls -la /ComfyUI/models/text_encoders/ 2>/dev/null || echo "(not found)"
echo ""
echo "Models in Docker image /ComfyUI/models/clip_vision/:"
ls -la /ComfyUI/models/clip_vision/ 2>/dev/null || echo "(not found)"
echo ""
echo "Models in Docker image /ComfyUI/models/loras/:"
ls -la /ComfyUI/models/loras/ 2>/dev/null || echo "(not found)"
# Copy models to network volume (if not already there)
echo ""
echo "=== Copying models to network volume ==="
# Diffusion models
if [ -d "/ComfyUI/models/diffusion_models" ]; then
echo "Copying diffusion models..."
cp -vn /ComfyUI/models/diffusion_models/*.safetensors /runpod-volume/models/diffusion_models/ 2>/dev/null || true
fi
# VAE models
if [ -d "/ComfyUI/models/vae" ]; then
echo "Copying VAE models..."
cp -vn /ComfyUI/models/vae/*.safetensors /runpod-volume/models/vae/ 2>/dev/null || true
fi
# Text encoders
if [ -d "/ComfyUI/models/text_encoders" ]; then
echo "Copying text encoder models..."
cp -vn /ComfyUI/models/text_encoders/*.safetensors /runpod-volume/models/text_encoders/ 2>/dev/null || true
fi
# CLIP vision
if [ -d "/ComfyUI/models/clip_vision" ]; then
echo "Copying CLIP vision models..."
cp -vn /ComfyUI/models/clip_vision/*.safetensors /runpod-volume/models/clip_vision/ 2>/dev/null || true
fi
# LoRAs
if [ -d "/ComfyUI/models/loras" ]; then
echo "Copying LoRA models..."
cp -vn /ComfyUI/models/loras/*.safetensors /runpod-volume/loras/ 2>/dev/null || true
fi
# Copy extra_model_paths.yaml to volume
echo ""
echo "Copying extra_model_paths.yaml to network volume..."
cat > /runpod-volume/extra_model_paths.yaml << 'EOF'
# ComfyUI Model Paths Configuration - Network Volume Priority
comfyui:
base_path: /ComfyUI/
is_default: true
checkpoints: |
/runpod-volume/models/checkpoints/
models/checkpoints/
clip: |
/runpod-volume/models/clip/
models/clip/
clip_vision: |
/runpod-volume/models/clip_vision/
models/clip_vision/
configs: models/configs/
controlnet: |
/runpod-volume/models/controlnet/
models/controlnet/
diffusion_models: |
/runpod-volume/models/diffusion_models/
/runpod-volume/models/
models/diffusion_models/
models/unet/
embeddings: |
/runpod-volume/models/embeddings/
models/embeddings/
loras: |
/runpod-volume/loras/
/runpod-volume/models/loras/
models/loras/
text_encoders: |
/runpod-volume/models/text_encoders/
models/text_encoders/
upscale_models: |
/runpod-volume/models/upscale_models/
models/upscale_models/
vae: |
/runpod-volume/models/vae/
models/vae/
EOF
echo ""
echo "=== Final network volume contents ==="
echo ""
echo "/runpod-volume/models/:"
du -sh /runpod-volume/models/*/ 2>/dev/null || echo "(empty)"
echo ""
echo "/runpod-volume/loras/:"
ls -la /runpod-volume/loras/ 2>/dev/null || echo "(empty)"
echo ""
echo "Total network volume usage:"
du -sh /runpod-volume/
echo ""
echo "=== Setup complete! ==="
echo "Models have been copied to the network volume."
echo "On subsequent cold starts, models will load from /runpod-volume/ (faster)."

249
scripts/worktree-manager.sh Executable file
View File

@ -0,0 +1,249 @@
#!/bin/bash
#
# Worktree Manager - Helper script for managing Git worktrees
#
set -e
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
REPO_ROOT="$(cd "$SCRIPT_DIR/.." && pwd)"
REPO_NAME=$(basename "$REPO_ROOT")
WORKTREE_BASE=$(dirname "$REPO_ROOT")
# Colors for output
RED='\033[0;31m'
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
BLUE='\033[0;34m'
CYAN='\033[0;36m'
NC='\033[0m' # No Color
show_help() {
cat << EOF
${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}
${GREEN}Worktree Manager${NC} - Manage Git worktrees easily
${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}
${YELLOW}Usage:${NC}
./worktree-manager.sh <command> [arguments]
${YELLOW}Commands:${NC}
${GREEN}list${NC} List all worktrees
${GREEN}create${NC} <branch> Create a new worktree for a branch
${GREEN}remove${NC} <branch> Remove a worktree
${GREEN}clean${NC} Remove all worktrees except main
${GREEN}goto${NC} <branch> Print command to cd to worktree
${GREEN}status${NC} Show status of all worktrees
${GREEN}help${NC} Show this help message
${YELLOW}Examples:${NC}
./worktree-manager.sh create feature/new-feature
./worktree-manager.sh list
./worktree-manager.sh remove feature/old-feature
./worktree-manager.sh clean
cd \$(./worktree-manager.sh goto feature/new-feature)
${YELLOW}Automatic Worktrees:${NC}
A Git hook is installed that automatically creates worktrees
when you run: ${CYAN}git checkout -b new-branch${NC}
${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}
EOF
}
list_worktrees() {
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
echo -e "${GREEN}Git Worktrees:${NC}"
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
cd "$REPO_ROOT"
git worktree list --porcelain | awk '
/^worktree/ { path=$2 }
/^HEAD/ { head=$2 }
/^branch/ {
branch=$2
gsub(/^refs\/heads\//, "", branch)
printf "%-40s %s\n", branch, path
}
/^detached/ {
printf "%-40s %s (detached)\n", head, path
}
' | while read line; do
if [[ $line == *"(detached)"* ]]; then
echo -e "${YELLOW} $line${NC}"
else
echo -e "${GREEN} $line${NC}"
fi
done
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
}
create_worktree() {
local branch=$1
if [ -z "$branch" ]; then
echo -e "${RED}Error: Branch name required${NC}"
echo "Usage: $0 create <branch-name>"
exit 1
fi
local worktree_path="${WORKTREE_BASE}/${REPO_NAME}-${branch}"
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
echo -e "${GREEN}Creating worktree for branch: ${YELLOW}$branch${NC}"
echo -e "${BLUE}Location: ${YELLOW}$worktree_path${NC}"
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
cd "$REPO_ROOT"
# Check if branch exists
if git show-ref --verify --quiet "refs/heads/$branch"; then
# Branch exists, just create worktree
git worktree add "$worktree_path" "$branch"
else
# Branch doesn't exist, create it
echo -e "${YELLOW}Branch doesn't exist, creating new branch...${NC}"
git worktree add -b "$branch" "$worktree_path"
fi
echo -e "${GREEN}✅ Worktree created successfully!${NC}"
echo -e ""
echo -e "To switch to the worktree:"
echo -e " ${CYAN}cd $worktree_path${NC}"
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
}
remove_worktree() {
local branch=$1
if [ -z "$branch" ]; then
echo -e "${RED}Error: Branch name required${NC}"
echo "Usage: $0 remove <branch-name>"
exit 1
fi
cd "$REPO_ROOT"
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
echo -e "${YELLOW}Removing worktree for branch: $branch${NC}"
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
git worktree remove "$branch" --force
echo -e "${GREEN}✅ Worktree removed successfully!${NC}"
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
}
clean_worktrees() {
cd "$REPO_ROOT"
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
echo -e "${YELLOW}Cleaning up worktrees (keeping main/master)...${NC}"
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
# Get list of worktrees excluding main/master
git worktree list --porcelain | grep "^branch" | sed 's/^branch refs\/heads\///' | while read branch; do
if [[ "$branch" != "main" ]] && [[ "$branch" != "master" ]]; then
echo -e "${YELLOW}Removing: $branch${NC}"
git worktree remove "$branch" --force 2>/dev/null || echo -e "${RED} Failed to remove $branch${NC}"
fi
done
# Prune deleted worktrees
git worktree prune
echo -e "${GREEN}✅ Cleanup complete!${NC}"
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
}
goto_worktree() {
local branch=$1
if [ -z "$branch" ]; then
echo -e "${RED}Error: Branch name required${NC}" >&2
exit 1
fi
cd "$REPO_ROOT"
# Find worktree path for branch
local worktree_path=$(git worktree list --porcelain | awk -v branch="$branch" '
/^worktree/ { path=$2 }
/^branch/ {
b=$2
gsub(/^refs\/heads\//, "", b)
if (b == branch) {
print path
exit
}
}
')
if [ -n "$worktree_path" ]; then
echo "$worktree_path"
else
echo -e "${RED}Error: No worktree found for branch '$branch'${NC}" >&2
exit 1
fi
}
show_status() {
cd "$REPO_ROOT"
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
echo -e "${GREEN}Worktree Status:${NC}"
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
git worktree list --porcelain | awk '
/^worktree/ { path=$2 }
/^branch/ {
branch=$2
gsub(/^refs\/heads\//, "", branch)
printf "\n%s%s%s\n", "Branch: ", branch, ""
printf "%s%s%s\n", "Path: ", path, ""
system("cd " path " && git status --short --branch | head -5")
}
' | while IFS= read -r line; do
if [[ $line == Branch:* ]]; then
echo -e "${GREEN}$line${NC}"
elif [[ $line == Path:* ]]; then
echo -e "${BLUE}$line${NC}"
else
echo -e "${YELLOW}$line${NC}"
fi
done
echo -e "${CYAN}━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
}
# Main command dispatcher
case "${1:-help}" in
list|ls)
list_worktrees
;;
create|add)
create_worktree "$2"
;;
remove|rm|delete)
remove_worktree "$2"
;;
clean|cleanup)
clean_worktrees
;;
goto|cd)
goto_worktree "$2"
;;
status|st)
show_status
;;
help|--help|-h)
show_help
;;
*)
echo -e "${RED}Unknown command: $1${NC}"
echo ""
show_help
exit 1
;;
esac

85
src/App.tsx
View File

@ -1,27 +1,20 @@
import "tldraw/tldraw.css"
import "@/css/style.css"
import { Default } from "@/routes/Default"
import { BrowserRouter, Route, Routes, Navigate } from "react-router-dom"
import { Contact } from "@/routes/Contact"
import { Board } from "./routes/Board"
import { Inbox } from "./routes/Inbox"
import { Presentations } from "./routes/Presentations"
import { Resilience } from "./routes/Resilience"
import { inject } from "@vercel/analytics"
import { createRoot } from "react-dom/client"
import { DailyProvider } from "@daily-co/daily-react"
import Daily from "@daily-co/daily-js"
import "tldraw/tldraw.css"; import "tldraw/tldraw.css";
import "@/css/style.css"; import "@/css/style.css";
import "@/css/auth.css"; // Import auth styles import "@/css/auth.css"; // Import auth styles
import "@/css/crypto-auth.css"; // Import crypto auth styles import "@/css/crypto-auth.css"; // Import crypto auth styles
import "@/css/starred-boards.css"; // Import starred boards styles import "@/css/starred-boards.css"; // Import starred boards styles
import "@/css/user-profile.css"; // Import user profile styles import "@/css/user-profile.css"; // Import user profile styles
import "@/css/location.css"; // Import location sharing styles import { Default } from "@/routes/Default";
import { BrowserRouter, Route, Routes, Navigate, useParams } from "react-router-dom";
import { Contact } from "@/routes/Contact";
import { Board } from "./routes/Board";
import { Inbox } from "./routes/Inbox";
import { Presentations } from "./routes/Presentations";
import { Resilience } from "./routes/Resilience";
import { Dashboard } from "./routes/Dashboard"; import { Dashboard } from "./routes/Dashboard";
import { LocationShareCreate } from "./routes/LocationShareCreate"; import { createRoot } from "react-dom/client";
import { LocationShareView } from "./routes/LocationShareView"; import { DailyProvider } from "@daily-co/daily-react";
import { LocationDashboardRoute } from "./routes/LocationDashboardRoute"; import Daily from "@daily-co/daily-js";
import { useState, useEffect } from 'react'; import { useState, useEffect } from 'react';
// Import React Context providers // Import React Context providers
@ -32,14 +25,12 @@ import NotificationsDisplay from './components/NotificationsDisplay';
import { ErrorBoundary } from './components/ErrorBoundary'; import { ErrorBoundary } from './components/ErrorBoundary';
// Import auth components // Import auth components
import CryptoLogin from './components/auth/CryptoLogin'; import CryptID from './components/auth/CryptID';
import CryptoDebug from './components/auth/CryptoDebug'; import CryptoDebug from './components/auth/CryptoDebug';
// Import Google Data test component // Import Google Data test component
import { GoogleDataTest } from './components/GoogleDataTest'; import { GoogleDataTest } from './components/GoogleDataTest';
inject();
// Initialize Daily.co call object with error handling // Initialize Daily.co call object with error handling
let callObject: any = null; let callObject: any = null;
try { try {
@ -77,6 +68,14 @@ const OptionalAuthRoute = ({ children }: { children: React.ReactNode }) => {
return <>{children}</>; return <>{children}</>;
}; };
/**
* Component to redirect board URLs without trailing slashes
*/
const RedirectBoardSlug = () => {
const { slug } = useParams<{ slug: string }>();
return <Navigate to={`/board/${slug}/`} replace />;
};
/** /**
* Main App with context providers * Main App with context providers
*/ */
@ -95,7 +94,7 @@ const AppWithProviders = () => {
return ( return (
<div className="auth-page"> <div className="auth-page">
<CryptoLogin onSuccess={() => window.location.href = '/'} /> <CryptID onSuccess={() => window.location.href = '/'} />
</div> </div>
); );
}; };
@ -111,66 +110,60 @@ const AppWithProviders = () => {
<NotificationsDisplay /> <NotificationsDisplay />
<Routes> <Routes>
{/* Redirect routes without trailing slashes to include them */}
<Route path="/login" element={<Navigate to="/login/" replace />} />
<Route path="/contact" element={<Navigate to="/contact/" replace />} />
<Route path="/board/:slug" element={<RedirectBoardSlug />} />
<Route path="/inbox" element={<Navigate to="/inbox/" replace />} />
<Route path="/debug" element={<Navigate to="/debug/" replace />} />
<Route path="/dashboard" element={<Navigate to="/dashboard/" replace />} />
<Route path="/presentations" element={<Navigate to="/presentations/" replace />} />
<Route path="/presentations/resilience" element={<Navigate to="/presentations/resilience/" replace />} />
{/* Auth routes */} {/* Auth routes */}
<Route path="/login" element={<AuthPage />} /> <Route path="/login/" element={<AuthPage />} />
{/* Optional auth routes */} {/* Optional auth routes */}
<Route path="/" element={ <Route path="/" element={
<OptionalAuthRoute> <OptionalAuthRoute>
<Default /> <Default />
</OptionalAuthRoute> </OptionalAuthRoute>
} /> } />
<Route path="/contact" element={ <Route path="/contact/" element={
<OptionalAuthRoute> <OptionalAuthRoute>
<Contact /> <Contact />
</OptionalAuthRoute> </OptionalAuthRoute>
} /> } />
<Route path="/board/:slug" element={ <Route path="/board/:slug/" element={
<OptionalAuthRoute> <OptionalAuthRoute>
<Board /> <Board />
</OptionalAuthRoute> </OptionalAuthRoute>
} /> } />
<Route path="/inbox" element={ <Route path="/inbox/" element={
<OptionalAuthRoute> <OptionalAuthRoute>
<Inbox /> <Inbox />
</OptionalAuthRoute> </OptionalAuthRoute>
} /> } />
<Route path="/debug" element={ <Route path="/debug/" element={
<OptionalAuthRoute> <OptionalAuthRoute>
<CryptoDebug /> <CryptoDebug />
</OptionalAuthRoute> </OptionalAuthRoute>
} /> } />
<Route path="/dashboard" element={ <Route path="/dashboard/" element={
<OptionalAuthRoute> <OptionalAuthRoute>
<Dashboard /> <Dashboard />
</OptionalAuthRoute> </OptionalAuthRoute>
} /> } />
<Route path="/presentations" element={ <Route path="/presentations/" element={
<OptionalAuthRoute> <OptionalAuthRoute>
<Presentations /> <Presentations />
</OptionalAuthRoute> </OptionalAuthRoute>
} /> } />
<Route path="/presentations/resilience" element={ <Route path="/presentations/resilience/" element={
<OptionalAuthRoute> <OptionalAuthRoute>
<Resilience /> <Resilience />
</OptionalAuthRoute> </OptionalAuthRoute>
} /> } />
{/* Location sharing routes */}
<Route path="/share-location" element={
<OptionalAuthRoute>
<LocationShareCreate />
</OptionalAuthRoute>
} />
<Route path="/location/:token" element={
<OptionalAuthRoute>
<LocationShareView />
</OptionalAuthRoute>
} />
<Route path="/location-dashboard" element={
<OptionalAuthRoute>
<LocationDashboardRoute />
</OptionalAuthRoute>
} />
{/* Google Data routes */} {/* Google Data routes */}
<Route path="/google" element={<GoogleDataTest />} /> <Route path="/google" element={<GoogleDataTest />} />
<Route path="/oauth/google/callback" element={<GoogleDataTest />} /> <Route path="/oauth/google/callback" element={<GoogleDataTest />} />

18
src/CmdK.tsx
View File

@ -66,11 +66,19 @@ export const CmdK = () => {
) )
const selected = editor.getSelectedShapeIds() const selected = editor.getSelectedShapeIds()
const inView = editor let inView: TLShapeId[] = []
.getShapesAtPoint(editor.getViewportPageBounds().center, { try {
margin: 1200, inView = editor
}) .getShapesAtPoint(editor.getViewportPageBounds().center, {
.map((o) => o.id) margin: 1200,
})
.map((o) => o.id)
} catch (e) {
// Some shapes may have invalid geometry (e.g., zero-length arrows)
// Fall back to getting all shapes on the current page
console.warn('getShapesAtPoint failed, falling back to all page shapes:', e)
inView = editor.getCurrentPageShapeIds() as unknown as TLShapeId[]
}
return new Map([ return new Map([
...nameToShapeIdMap, ...nameToShapeIdMap,

571
src/automerge/AutomergeToTLStore.ts
View File

@ -1,9 +1,57 @@
import { TLRecord, RecordId, TLStore } from "@tldraw/tldraw" import { TLRecord, RecordId, TLStore, IndexKey } from "@tldraw/tldraw"
import * as Automerge from "@automerge/automerge" import * as Automerge from "@automerge/automerge"
// Helper function to validate if a string is a valid tldraw IndexKey
// tldraw uses fractional indexing based on https://observablehq.com/@dgreensp/implementing-fractional-indexing
// The first letter encodes integer part length: a=1 digit, b=2 digits, c=3 digits, etc.
// Examples: "a0"-"a9", "b10"-"b99", "c100"-"c999", with optional fraction "a1V4rr"
// Invalid: "b1" (b expects 2 digits but has 1), simple sequential numbers
function isValidIndexKey(index: string): boolean {
if (!index || typeof index !== 'string' || index.length === 0) {
return false
}
// Must start with a letter
if (!/^[a-zA-Z]/.test(index)) {
return false
}
const prefix = index[0]
const rest = index.slice(1)
// For lowercase prefixes, validate digit count matches the prefix
if (prefix >= 'a' && prefix <= 'z') {
// Calculate expected minimum digit count: a=1, b=2, c=3, etc.
const expectedDigits = prefix.charCodeAt(0) - 'a'.charCodeAt(0) + 1
// Extract the integer part (leading digits)
const integerMatch = rest.match(/^(\d+)/)
if (!integerMatch) {
// No digits at all - invalid
return false
}
const integerPart = integerMatch[1]
// Check if integer part has correct number of digits for the prefix
if (integerPart.length < expectedDigits) {
// Invalid: "b1" has b (expects 2 digits) but only has 1 digit
return false
}
}
// Check overall format: letter followed by alphanumeric
if (/^[a-zA-Z][a-zA-Z0-9]+$/.test(index)) {
return true
}
return false
}
export function applyAutomergePatchesToTLStore( export function applyAutomergePatchesToTLStore(
patches: Automerge.Patch[], patches: Automerge.Patch[],
store: TLStore store: TLStore,
automergeDoc?: any // Optional Automerge document to read full records from
) { ) {
const toRemove: TLRecord["id"][] = [] const toRemove: TLRecord["id"][] = []
const updatedObjects: { [id: string]: TLRecord } = {} const updatedObjects: { [id: string]: TLRecord } = {}
@ -27,10 +75,11 @@ export function applyAutomergePatchesToTLStore(
const existingRecord = getRecordFromStore(store, id) const existingRecord = getRecordFromStore(store, id)
// CRITICAL: For shapes, get coordinates from store's current state BEFORE any patch processing // CRITICAL: For shapes, get coordinates and parentId from store's current state BEFORE any patch processing
// This ensures we preserve coordinates even if patches don't include them // This ensures we preserve coordinates and parent relationships even if patches don't include them
// This is especially important when patches come back after store.put operations // This is especially important when patches come back after store.put operations
let storeCoordinates: { x?: number; y?: number } = {} let storeCoordinates: { x?: number; y?: number } = {}
let storeParentId: string | undefined = undefined
if (existingRecord && existingRecord.typeName === 'shape') { if (existingRecord && existingRecord.typeName === 'shape') {
const storeX = (existingRecord as any).x const storeX = (existingRecord as any).x
const storeY = (existingRecord as any).y const storeY = (existingRecord as any).y
@ -40,6 +89,39 @@ export function applyAutomergePatchesToTLStore(
if (typeof storeY === 'number' && !isNaN(storeY) && storeY !== null && storeY !== undefined) { if (typeof storeY === 'number' && !isNaN(storeY) && storeY !== null && storeY !== undefined) {
storeCoordinates.y = storeY storeCoordinates.y = storeY
} }
// CRITICAL: Preserve parentId from store (might be a frame or group!)
const existingParentId = (existingRecord as any).parentId
if (existingParentId && typeof existingParentId === 'string') {
storeParentId = existingParentId
}
}
// CRITICAL: If record doesn't exist in store yet, try to get it from Automerge document
// This prevents coordinates from defaulting to 0,0 when patches create new records
let automergeRecord: any = null
let automergeParentId: string | undefined = undefined
if (!existingRecord && automergeDoc && automergeDoc.store && automergeDoc.store[id]) {
try {
automergeRecord = automergeDoc.store[id]
// Extract coordinates and parentId from Automerge record if it's a shape
if (automergeRecord && automergeRecord.typeName === 'shape') {
const docX = automergeRecord.x
const docY = automergeRecord.y
if (typeof docX === 'number' && !isNaN(docX) && docX !== null && docX !== undefined) {
storeCoordinates.x = docX
}
if (typeof docY === 'number' && !isNaN(docY) && docY !== null && docY !== undefined) {
storeCoordinates.y = docY
}
// CRITICAL: Preserve parentId from Automerge document (might be a frame!)
if (automergeRecord.parentId && typeof automergeRecord.parentId === 'string') {
automergeParentId = automergeRecord.parentId
}
}
} catch (e) {
// If we can't read from Automerge doc, continue without it
console.warn(`Could not read record ${id} from Automerge document:`, e)
}
} }
// Infer typeName from ID pattern if record doesn't exist // Infer typeName from ID pattern if record doesn't exist
@ -112,7 +194,20 @@ export function applyAutomergePatchesToTLStore(
} }
} }
let record = updatedObjects[id] || (existingRecord ? JSON.parse(JSON.stringify(existingRecord)) : defaultRecord) // CRITICAL: When creating a new record, prefer using the full record from Automerge document
// This ensures we get all properties including coordinates, not just defaults
let record: any
if (updatedObjects[id]) {
record = updatedObjects[id]
} else if (existingRecord) {
record = JSON.parse(JSON.stringify(existingRecord))
} else if (automergeRecord) {
// Use the full record from Automerge document - this has all properties including coordinates
record = JSON.parse(JSON.stringify(automergeRecord))
} else {
// Fallback to default record only if we can't get it from anywhere else
record = defaultRecord
}
// CRITICAL: For shapes, ensure x and y are always present (even if record came from updatedObjects) // CRITICAL: For shapes, ensure x and y are always present (even if record came from updatedObjects)
// This prevents coordinates from being lost when records are created from patches // This prevents coordinates from being lost when records are created from patches
@ -157,6 +252,28 @@ export function applyAutomergePatchesToTLStore(
const originalX = storeCoordinates.x !== undefined ? storeCoordinates.x : recordX const originalX = storeCoordinates.x !== undefined ? storeCoordinates.x : recordX
const originalY = storeCoordinates.y !== undefined ? storeCoordinates.y : recordY const originalY = storeCoordinates.y !== undefined ? storeCoordinates.y : recordY
const hadOriginalCoordinates = originalX !== undefined && originalY !== undefined const hadOriginalCoordinates = originalX !== undefined && originalY !== undefined
// CRITICAL: Store original richText and arrow text before patch application to preserve them
// This ensures richText and arrow text aren't lost when patches only update other properties
let originalRichText: any = undefined
let originalArrowText: any = undefined
if (record.typeName === 'shape') {
// Get richText from store's current state (most reliable)
if (existingRecord && (existingRecord as any).props && (existingRecord as any).props.richText) {
originalRichText = (existingRecord as any).props.richText
} else if ((record as any).props && (record as any).props.richText) {
originalRichText = (record as any).props.richText
}
// Get arrow text from store's current state (most reliable)
if ((record as any).type === 'arrow') {
if (existingRecord && (existingRecord as any).props && (existingRecord as any).props.text !== undefined) {
originalArrowText = (existingRecord as any).props.text
} else if ((record as any).props && (record as any).props.text !== undefined) {
originalArrowText = (record as any).props.text
}
}
}
switch (patch.action) { switch (patch.action) {
case "insert": { case "insert": {
@ -229,6 +346,58 @@ export function applyAutomergePatchesToTLStore(
updatedObjects[id] = { ...updatedObjects[id], y: defaultRecord.y || 0 } as TLRecord updatedObjects[id] = { ...updatedObjects[id], y: defaultRecord.y || 0 } as TLRecord
} }
} }
// CRITICAL: Preserve richText and arrow text after patch application
// This prevents richText and arrow text from being lost when patches only update other properties
const currentRecord = updatedObjects[id]
// Preserve richText for geo/note/text shapes
if (originalRichText !== undefined && (currentRecord as any).type !== 'arrow') {
const patchedProps = (currentRecord as any).props || {}
const patchedRichText = patchedProps.richText
// If patch didn't include richText, preserve the original
if (patchedRichText === undefined || patchedRichText === null) {
updatedObjects[id] = {
...currentRecord,
props: {
...patchedProps,
richText: originalRichText
}
} as TLRecord
}
}
// Preserve arrow text for arrow shapes
if (originalArrowText !== undefined && (currentRecord as any).type === 'arrow') {
const patchedProps = (currentRecord as any).props || {}
const patchedText = patchedProps.text
// If patch didn't include text, preserve the original
if (patchedText === undefined || patchedText === null) {
updatedObjects[id] = {
...currentRecord,
props: {
...patchedProps,
text: originalArrowText
}
} as TLRecord
}
}
// CRITICAL: Preserve parentId from store or Automerge document
// This prevents shapes from losing their frame/group parent relationships
// which causes them to reset to (0,0) on the page instead of maintaining their position in the frame
// Priority: store parentId (most reliable), then Automerge parentId, then patch value
const preservedParentId = storeParentId || automergeParentId
if (preservedParentId !== undefined) {
const patchedParentId = (currentRecord as any).parentId
// If patch didn't include parentId, or it's missing/default, use the preserved parentId
if (!patchedParentId || (patchedParentId === 'page:page' && preservedParentId !== 'page:page')) {
updatedObjects[id] = {
...currentRecord,
parentId: preservedParentId
} as TLRecord
}
}
} }
// CRITICAL: Re-check typeName after patch application to ensure it's still correct // CRITICAL: Re-check typeName after patch application to ensure it's still correct
@ -251,6 +420,12 @@ export function applyAutomergePatchesToTLStore(
return // Skip - not a TLDraw record return // Skip - not a TLDraw record
} }
// Filter out SharedPiano shapes since they're no longer supported
if (record.typeName === 'shape' && (record as any).type === 'SharedPiano') {
console.log(`⚠️ Filtering out deprecated SharedPiano shape: ${record.id}`)
return // Skip - SharedPiano is deprecated
}
try { try {
const sanitized = sanitizeRecord(record) const sanitized = sanitizeRecord(record)
toPut.push(sanitized) toPut.push(sanitized)
@ -270,6 +445,18 @@ export function applyAutomergePatchesToTLStore(
// put / remove the records in the store // put / remove the records in the store
// Log patch application for debugging // Log patch application for debugging
console.log(`🔧 AutomergeToTLStore: Applying ${patches.length} patches, ${toPut.length} records to put, ${toRemove.length} records to remove`) console.log(`🔧 AutomergeToTLStore: Applying ${patches.length} patches, ${toPut.length} records to put, ${toRemove.length} records to remove`)
// DEBUG: Log shape updates being applied to store
toPut.forEach(record => {
if (record.typeName === 'shape' && (record as any).props?.w) {
console.log(`🔧 AutomergeToTLStore: Putting shape ${(record as any).type} ${record.id}:`, {
w: (record as any).props.w,
h: (record as any).props.h,
x: (record as any).x,
y: (record as any).y
})
}
})
if (failedRecords.length > 0) { if (failedRecords.length > 0) {
console.log({ patches, toPut: toPut.length, failed: failedRecords.length }) console.log({ patches, toPut: toPut.length, failed: failedRecords.length })
@ -402,14 +589,25 @@ export function sanitizeRecord(record: any): TLRecord {
// For shapes, only ensure basic required fields exist // For shapes, only ensure basic required fields exist
if (sanitized.typeName === 'shape') { if (sanitized.typeName === 'shape') {
// CRITICAL: Remove instance-only properties from shapes (these cause validation errors)
// These properties should only exist on instance records, not shape records
const instanceOnlyProperties = ['insets', 'brush', 'zoomBrush', 'scribbles', 'duplicateProps']
instanceOnlyProperties.forEach(prop => {
if (prop in sanitized) {
delete (sanitized as any)[prop]
}
})
// Ensure required shape fields exist // Ensure required shape fields exist
// CRITICAL: Only set defaults if coordinates are truly missing or invalid // CRITICAL: Only set defaults if coordinates are truly missing or invalid
// DO NOT overwrite valid coordinates (including 0, which is a valid position) // DO NOT overwrite valid coordinates (including 0, which is a valid position)
// Only set to 0 if the value is undefined, null, or NaN // Only set to 0 if the value is undefined, null, or NaN
if (sanitized.x === undefined || sanitized.x === null || (typeof sanitized.x === 'number' && isNaN(sanitized.x))) { if (sanitized.x === undefined || sanitized.x === null || (typeof sanitized.x === 'number' && isNaN(sanitized.x))) {
console.warn(`⚠️ Shape ${sanitized.id} (${sanitized.type}) has invalid x coordinate, defaulting to 0. Original value:`, sanitized.x)
sanitized.x = 0 sanitized.x = 0
} }
if (sanitized.y === undefined || sanitized.y === null || (typeof sanitized.y === 'number' && isNaN(sanitized.y))) { if (sanitized.y === undefined || sanitized.y === null || (typeof sanitized.y === 'number' && isNaN(sanitized.y))) {
console.warn(`⚠️ Shape ${sanitized.id} (${sanitized.type}) has invalid y coordinate, defaulting to 0. Original value:`, sanitized.y)
sanitized.y = 0 sanitized.y = 0
} }
if (typeof sanitized.rotation !== 'number') sanitized.rotation = 0 if (typeof sanitized.rotation !== 'number') sanitized.rotation = 0
@ -422,21 +620,151 @@ export function sanitizeRecord(record: any): TLRecord {
// Ensure meta is a mutable copy to preserve all properties (including text for rectangles) // Ensure meta is a mutable copy to preserve all properties (including text for rectangles)
sanitized.meta = { ...sanitized.meta } sanitized.meta = { ...sanitized.meta }
} }
if (!sanitized.index) sanitized.index = 'a1' // CRITICAL: IndexKey must follow tldraw's fractional indexing format
// Valid format: starts with 'a' followed by digits, optionally followed by alphanumeric jitter
// Examples: "a1", "a2", "a10", "a1V", "a24sT", "a1V4rr" (fractional between a1 and a2)
// Invalid: "c1", "b1", "z999" (old format - not valid fractional indices)
if (!isValidIndexKey(sanitized.index)) {
console.warn(`⚠️ Invalid index "${sanitized.index}" for shape ${sanitized.id}, resetting to 'a1'`)
sanitized.index = 'a1' as IndexKey
}
if (!sanitized.parentId) sanitized.parentId = 'page:page' if (!sanitized.parentId) sanitized.parentId = 'page:page'
if (!sanitized.props || typeof sanitized.props !== 'object') sanitized.props = {} if (!sanitized.props || typeof sanitized.props !== 'object') sanitized.props = {}
// CRITICAL: Ensure props is a deep mutable copy to preserve all nested properties // CRITICAL: Ensure props is a deep mutable copy to preserve all nested properties
// This is essential for custom shapes like ObsNote and for preserving richText in geo shapes // This is essential for custom shapes like ObsNote and for preserving richText in geo shapes
// Use JSON parse/stringify to create a deep copy of nested objects (like richText.content) // Use JSON parse/stringify to create a deep copy of nested objects (like richText.content)
sanitized.props = JSON.parse(JSON.stringify(sanitized.props)) try {
sanitized.props = JSON.parse(JSON.stringify(sanitized.props))
} catch (e) {
// If JSON serialization fails (e.g., due to functions or circular references),
// create a shallow copy and recursively clean it
console.warn(`⚠️ Could not deep copy props for shape ${sanitized.id}, using shallow copy:`, e)
const propsCopy: any = {}
for (const key in sanitized.props) {
try {
const value = sanitized.props[key]
// Skip functions
if (typeof value === 'function') {
continue
}
// Try to serialize individual values
try {
propsCopy[key] = JSON.parse(JSON.stringify(value))
} catch (valueError) {
// If individual value can't be serialized, use it as-is if it's a primitive
if (value === null || value === undefined || typeof value !== 'object') {
propsCopy[key] = value
}
// Otherwise skip it
}
} catch (keyError) {
// Skip properties that can't be accessed
continue
}
}
sanitized.props = propsCopy
}
// CRITICAL: Map old shape type names to new ones (migration support) // CRITICAL: Map old shape type names to new ones (migration support)
// This handles renamed shape types from old data // This handles renamed shape types from old data
if (sanitized.type === 'Transcribe') { if (sanitized.type === 'Transcribe') {
sanitized.type = 'Transcription' sanitized.type = 'Transcription'
} }
// CRITICAL: Normalize case for custom shape types (lowercase → PascalCase)
// The schema expects PascalCase (e.g., "ChatBox" not "chatBox")
const customShapeTypeMap: Record<string, string> = {
'chatBox': 'ChatBox',
'videoChat': 'VideoChat',
'embed': 'Embed',
'markdown': 'Markdown',
'mycrozineTemplate': 'MycrozineTemplate',
'slide': 'Slide',
'prompt': 'Prompt',
'transcription': 'Transcription',
'obsNote': 'ObsNote',
'fathomNote': 'FathomNote',
'holon': 'Holon',
'obsidianBrowser': 'ObsidianBrowser',
'fathomMeetingsBrowser': 'FathomMeetingsBrowser',
'imageGen': 'ImageGen',
'videoGen': 'VideoGen',
'multmux': 'Multmux',
}
// Normalize the shape type if it's a custom type with incorrect case
if (sanitized.type && typeof sanitized.type === 'string' && customShapeTypeMap[sanitized.type]) {
console.log(`🔧 Normalizing shape type: "${sanitized.type}" → "${customShapeTypeMap[sanitized.type]}"`)
sanitized.type = customShapeTypeMap[sanitized.type]
}
// CRITICAL: Sanitize Multmux shapes AFTER case normalization - ensure all required props exist
// Old shapes may have wsUrl (removed) or undefined values
if (sanitized.type === 'Multmux') {
console.log(`🔧 Sanitizing Multmux shape ${sanitized.id}:`, JSON.stringify(sanitized.props))
// Remove deprecated wsUrl prop
if ('wsUrl' in sanitized.props) {
delete sanitized.props.wsUrl
}
// CRITICAL: Create a clean props object with all required values
// This ensures no undefined values slip through validation
// Every value MUST be explicitly defined - undefined values cause ValidationError
const w = (typeof sanitized.props.w === 'number' && !isNaN(sanitized.props.w)) ? sanitized.props.w : 800
const h = (typeof sanitized.props.h === 'number' && !isNaN(sanitized.props.h)) ? sanitized.props.h : 600
const sessionId = (typeof sanitized.props.sessionId === 'string') ? sanitized.props.sessionId : ''
const sessionName = (typeof sanitized.props.sessionName === 'string') ? sanitized.props.sessionName : ''
const token = (typeof sanitized.props.token === 'string') ? sanitized.props.token : ''
// Fix old port (3000 -> 3002) during sanitization
let serverUrl = (typeof sanitized.props.serverUrl === 'string') ? sanitized.props.serverUrl : 'http://localhost:3002'
if (serverUrl === 'http://localhost:3000') {
serverUrl = 'http://localhost:3002'
}
const pinnedToView = (sanitized.props.pinnedToView === true) ? true : false
// Filter out any undefined or non-string elements from tags array
let tags: string[] = ['terminal', 'multmux']
if (Array.isArray(sanitized.props.tags)) {
const filteredTags = sanitized.props.tags.filter((t: any) => typeof t === 'string' && t !== '')
if (filteredTags.length > 0) {
tags = filteredTags
}
}
// Build clean props object - all values are guaranteed to be defined
const cleanProps = {
w: w,
h: h,
sessionId: sessionId,
sessionName: sessionName,
token: token,
serverUrl: serverUrl,
pinnedToView: pinnedToView,
tags: tags,
}
// CRITICAL: Verify no undefined values before assigning
// This is a safety check - if any value is undefined, something went wrong above
for (const [key, value] of Object.entries(cleanProps)) {
if (value === undefined) {
console.error(`❌ CRITICAL: Multmux prop ${key} is undefined after sanitization! This should never happen.`)
// Fix it with a default value based on key
switch (key) {
case 'w': (cleanProps as any).w = 800; break
case 'h': (cleanProps as any).h = 600; break
case 'sessionId': (cleanProps as any).sessionId = ''; break
case 'sessionName': (cleanProps as any).sessionName = ''; break
case 'token': (cleanProps as any).token = ''; break
case 'serverUrl': (cleanProps as any).serverUrl = 'http://localhost:3002'; break
case 'pinnedToView': (cleanProps as any).pinnedToView = false; break
case 'tags': (cleanProps as any).tags = ['terminal', 'multmux']; break
}
}
}
sanitized.props = cleanProps
console.log(`🔧 Sanitized Multmux shape ${sanitized.id} props:`, JSON.stringify(sanitized.props))
}
// CRITICAL: Infer type from properties BEFORE defaulting to 'geo' // CRITICAL: Infer type from properties BEFORE defaulting to 'geo'
// This ensures arrows and other shapes are properly recognized // This ensures arrows and other shapes are properly recognized
if (!sanitized.type || typeof sanitized.type !== 'string') { if (!sanitized.type || typeof sanitized.type !== 'string') {
@ -571,15 +899,63 @@ export function sanitizeRecord(record: any): TLRecord {
// Remove invalid w/h from props (they cause validation errors) // Remove invalid w/h from props (they cause validation errors)
if ('w' in sanitized.props) delete sanitized.props.w if ('w' in sanitized.props) delete sanitized.props.w
if ('h' in sanitized.props) delete sanitized.props.h if ('h' in sanitized.props) delete sanitized.props.h
// Line shapes REQUIRE points property // Line shapes REQUIRE points property with at least 2 points
if (!sanitized.props.points || typeof sanitized.props.points !== 'object' || Array.isArray(sanitized.props.points)) { if (!sanitized.props.points || typeof sanitized.props.points !== 'object' || Array.isArray(sanitized.props.points)) {
sanitized.props.points = { sanitized.props.points = {
'a1': { id: 'a1', index: 'a1' as any, x: 0, y: 0 }, 'a1': { id: 'a1', index: 'a1' as any, x: 0, y: 0 },
'a2': { id: 'a2', index: 'a2' as any, x: 100, y: 0 } 'a2': { id: 'a2', index: 'a2' as any, x: 100, y: 0 }
} }
} else {
// Ensure the points object has at least 2 valid points
const pointKeys = Object.keys(sanitized.props.points)
if (pointKeys.length < 2) {
sanitized.props.points = {
'a1': { id: 'a1', index: 'a1' as any, x: 0, y: 0 },
'a2': { id: 'a2', index: 'a2' as any, x: 100, y: 0 }
}
}
} }
} }
// CRITICAL: Fix draw shapes - ensure valid segments structure (required by schema)
// Draw shapes with empty segments cause "No nearest point found" errors
if (sanitized.type === 'draw') {
// Remove invalid w/h from props (they cause validation errors)
if ('w' in sanitized.props) delete sanitized.props.w
if ('h' in sanitized.props) delete sanitized.props.h
// Draw shapes REQUIRE segments property with at least one segment containing points
if (!sanitized.props.segments || !Array.isArray(sanitized.props.segments) || sanitized.props.segments.length === 0) {
// Create a minimal valid segment with at least 2 points
sanitized.props.segments = [{
type: 'free',
points: [
{ x: 0, y: 0, z: 0.5 },
{ x: 10, y: 0, z: 0.5 }
]
}]
} else {
// Ensure each segment has valid points
sanitized.props.segments = sanitized.props.segments.map((segment: any) => {
if (!segment.points || !Array.isArray(segment.points) || segment.points.length < 2) {
return {
type: segment.type || 'free',
points: [
{ x: 0, y: 0, z: 0.5 },
{ x: 10, y: 0, z: 0.5 }
]
}
}
return segment
})
}
// Ensure required draw shape properties exist
if (typeof sanitized.props.isClosed !== 'boolean') sanitized.props.isClosed = false
if (typeof sanitized.props.isComplete !== 'boolean') sanitized.props.isComplete = true
if (typeof sanitized.props.isPen !== 'boolean') sanitized.props.isPen = false
}
// CRITICAL: Fix group shapes - remove invalid w/h from props // CRITICAL: Fix group shapes - remove invalid w/h from props
if (sanitized.type === 'group') { if (sanitized.type === 'group') {
@ -603,6 +979,33 @@ export function sanitizeRecord(record: any): TLRecord {
} }
} }
// CRITICAL: Convert props.text to props.richText for geo shapes (tldraw schema change)
// tldraw no longer accepts props.text on geo shapes - must use richText
// This migration handles shapes that were saved before the schema change
if (sanitized.type === 'geo' && 'text' in sanitized.props && typeof sanitized.props.text === 'string') {
const textContent = sanitized.props.text
// Convert text string to richText format for tldraw
sanitized.props.richText = {
type: 'doc',
content: textContent ? [{
type: 'paragraph',
content: [{
type: 'text',
text: textContent
}]
}] : []
}
// CRITICAL: Preserve original text in meta.text for backward compatibility
// This is used by search (src/utils/searchUtils.ts) and other legacy code
if (!sanitized.meta) sanitized.meta = {}
sanitized.meta.text = textContent
// Remove invalid props.text
delete sanitized.props.text
}
// CRITICAL: Fix richText structure for geo shapes (preserve content) // CRITICAL: Fix richText structure for geo shapes (preserve content)
if (sanitized.type === 'geo' && sanitized.props.richText) { if (sanitized.type === 'geo' && sanitized.props.richText) {
if (Array.isArray(sanitized.props.richText)) { if (Array.isArray(sanitized.props.richText)) {
@ -615,6 +1018,96 @@ export function sanitizeRecord(record: any): TLRecord {
sanitized.props.richText = cleanRichTextNaN(sanitized.props.richText) sanitized.props.richText = cleanRichTextNaN(sanitized.props.richText)
} }
// CRITICAL: Fix arrow shapes - ensure valid start/end structure (required by schema)
// Arrows with invalid start/end cause "No nearest point found" errors
if (sanitized.type === 'arrow') {
// Ensure start property exists and has valid structure
if (!sanitized.props.start || typeof sanitized.props.start !== 'object') {
sanitized.props.start = { x: 0, y: 0 }
} else {
// Ensure start has x and y properties (could be bound to a shape or free)
const start = sanitized.props.start as any
if (start.type === 'binding') {
// Binding type must have boundShapeId, normalizedAnchor, and other properties
if (!start.boundShapeId) {
// Invalid binding - convert to point
sanitized.props.start = { x: start.x ?? 0, y: start.y ?? 0 }
}
} else if (start.type === 'point' || start.type === undefined) {
// Point type must have x and y
if (typeof start.x !== 'number' || typeof start.y !== 'number') {
sanitized.props.start = { x: 0, y: 0 }
}
}
}
// Ensure end property exists and has valid structure
if (!sanitized.props.end || typeof sanitized.props.end !== 'object') {
sanitized.props.end = { x: 100, y: 0 }
} else {
// Ensure end has x and y properties (could be bound to a shape or free)
const end = sanitized.props.end as any
if (end.type === 'binding') {
// Binding type must have boundShapeId
if (!end.boundShapeId) {
// Invalid binding - convert to point
sanitized.props.end = { x: end.x ?? 100, y: end.y ?? 0 }
}
} else if (end.type === 'point' || end.type === undefined) {
// Point type must have x and y
if (typeof end.x !== 'number' || typeof end.y !== 'number') {
sanitized.props.end = { x: 100, y: 0 }
}
}
}
// Ensure bend is a valid number
if (typeof sanitized.props.bend !== 'number' || isNaN(sanitized.props.bend)) {
sanitized.props.bend = 0
}
// Ensure arrowhead properties exist
if (!sanitized.props.arrowheadStart) sanitized.props.arrowheadStart = 'none'
if (!sanitized.props.arrowheadEnd) sanitized.props.arrowheadEnd = 'arrow'
// Ensure text property exists and is a string
if (sanitized.props.text === undefined || sanitized.props.text === null) {
sanitized.props.text = ''
} else if (typeof sanitized.props.text !== 'string') {
// If text is not a string (e.g., RichText object), convert it to string
try {
if (typeof sanitized.props.text === 'object' && sanitized.props.text !== null) {
// Try to extract text from RichText object
const textObj = sanitized.props.text as any
if (Array.isArray(textObj.content)) {
// Extract text from RichText content
const extractText = (content: any[]): string => {
return content.map((item: any) => {
if (item.type === 'text' && item.text) {
return item.text
} else if (item.content && Array.isArray(item.content)) {
return extractText(item.content)
}
return ''
}).join('')
}
sanitized.props.text = extractText(textObj.content)
} else if (textObj.text && typeof textObj.text === 'string') {
sanitized.props.text = textObj.text
} else {
sanitized.props.text = String(sanitized.props.text)
}
} else {
sanitized.props.text = String(sanitized.props.text)
}
} catch (e) {
console.warn(`⚠️ AutomergeToTLStore: Error converting arrow text to string for ${sanitized.id}:`, e)
sanitized.props.text = String(sanitized.props.text)
}
}
// Note: We preserve text even if it's an empty string - that's a valid value
}
// CRITICAL: Fix richText structure for text shapes - REQUIRED field // CRITICAL: Fix richText structure for text shapes - REQUIRED field
if (sanitized.type === 'text') { if (sanitized.type === 'text') {
// Text shapes MUST have props.richText as an object - initialize if missing // Text shapes MUST have props.richText as an object - initialize if missing
@ -628,10 +1121,35 @@ export function sanitizeRecord(record: any): TLRecord {
} }
// CRITICAL: Clean NaN values from richText content to prevent SVG export errors // CRITICAL: Clean NaN values from richText content to prevent SVG export errors
sanitized.props.richText = cleanRichTextNaN(sanitized.props.richText) sanitized.props.richText = cleanRichTextNaN(sanitized.props.richText)
// CRITICAL: Ensure required text shape properties exist (TLDraw validation requires these)
// color is REQUIRED and must be one of the valid color values
const validColors = ['black', 'grey', 'light-violet', 'violet', 'blue', 'light-blue', 'yellow', 'orange', 'green', 'light-green', 'light-red', 'red', 'white']
if (!sanitized.props.color || typeof sanitized.props.color !== 'string' || !validColors.includes(sanitized.props.color)) {
sanitized.props.color = 'black'
}
// Ensure other required properties have defaults
if (typeof sanitized.props.w !== 'number') sanitized.props.w = 300
if (!sanitized.props.size || typeof sanitized.props.size !== 'string') sanitized.props.size = 'm'
if (!sanitized.props.font || typeof sanitized.props.font !== 'string') sanitized.props.font = 'draw'
if (!sanitized.props.textAlign || typeof sanitized.props.textAlign !== 'string') sanitized.props.textAlign = 'start'
if (typeof sanitized.props.autoSize !== 'boolean') sanitized.props.autoSize = false
if (typeof sanitized.props.scale !== 'number') sanitized.props.scale = 1
// Remove invalid properties for text shapes (these cause validation errors)
// Remove properties that are only valid for custom shapes, not standard TLDraw text shapes
// CRITICAL: 'text' property is NOT allowed - text shapes must use props.richText instead
const invalidTextProps = ['h', 'geo', 'text', 'isEditing', 'editingContent', 'isTranscribing', 'isPaused', 'fixedHeight', 'pinnedToView', 'isModified', 'originalContent', 'editingName', 'editingDescription', 'isConnected', 'holonId', 'noteId', 'title', 'content', 'tags', 'showPreview', 'backgroundColor', 'textColor']
invalidTextProps.forEach(prop => {
if (prop in sanitized.props) {
delete sanitized.props[prop]
}
})
} }
// CRITICAL: Remove invalid 'text' property from text shapes (TLDraw schema doesn't allow props.text) // CRITICAL: Additional safety check - Remove invalid 'text' property from text shapes
// Text shapes should only use props.richText, not props.text // Text shapes should only use props.richText, not props.text
// This is a redundant check to ensure text property is always removed
if (sanitized.type === 'text' && 'text' in sanitized.props) { if (sanitized.type === 'text' && 'text' in sanitized.props) {
delete sanitized.props.text delete sanitized.props.text
} }
@ -655,9 +1173,28 @@ export function sanitizeRecord(record: any): TLRecord {
// CRITICAL: Clean NaN values from richText content to prevent SVG export errors // CRITICAL: Clean NaN values from richText content to prevent SVG export errors
sanitized.props.richText = cleanRichTextNaN(sanitized.props.richText) sanitized.props.richText = cleanRichTextNaN(sanitized.props.richText)
// Only remove properties that cause validation errors (not all "invalid" ones) // CRITICAL: Ensure required text shape properties exist (TLDraw validation requires these)
if ('h' in sanitized.props) delete sanitized.props.h // color is REQUIRED and must be one of the valid color values
if ('geo' in sanitized.props) delete sanitized.props.geo const validColors = ['black', 'grey', 'light-violet', 'violet', 'blue', 'light-blue', 'yellow', 'orange', 'green', 'light-green', 'light-red', 'red', 'white']
if (!sanitized.props.color || typeof sanitized.props.color !== 'string' || !validColors.includes(sanitized.props.color)) {
sanitized.props.color = 'black'
}
// Ensure other required properties have defaults
if (typeof sanitized.props.w !== 'number') sanitized.props.w = 300
if (!sanitized.props.size || typeof sanitized.props.size !== 'string') sanitized.props.size = 'm'
if (!sanitized.props.font || typeof sanitized.props.font !== 'string') sanitized.props.font = 'draw'
if (!sanitized.props.textAlign || typeof sanitized.props.textAlign !== 'string') sanitized.props.textAlign = 'start'
if (typeof sanitized.props.autoSize !== 'boolean') sanitized.props.autoSize = false
if (typeof sanitized.props.scale !== 'number') sanitized.props.scale = 1
// Remove invalid properties for text shapes (these cause validation errors)
// Remove properties that are only valid for custom shapes, not standard TLDraw text shapes
const invalidTextProps = ['h', 'geo', 'isEditing', 'editingContent', 'isTranscribing', 'isPaused', 'fixedHeight', 'pinnedToView', 'isModified', 'originalContent', 'editingName', 'editingDescription', 'isConnected', 'holonId', 'noteId', 'title', 'content', 'tags', 'showPreview', 'backgroundColor', 'textColor']
invalidTextProps.forEach(prop => {
if (prop in sanitized.props) {
delete sanitized.props[prop]
}
})
} }
} else if (sanitized.typeName === 'instance') { } else if (sanitized.typeName === 'instance') {
// CRITICAL: Handle instance records - ensure required fields exist // CRITICAL: Handle instance records - ensure required fields exist
@ -702,6 +1239,12 @@ export function sanitizeRecord(record: any): TLRecord {
} }
} }
// CRITICAL: Final safety check - ensure text shapes never have invalid 'text' property
// This is a last-resort check before returning to catch any edge cases
if (sanitized.typeName === 'shape' && sanitized.type === 'text' && sanitized.props && 'text' in sanitized.props) {
delete sanitized.props.text
}
return sanitized return sanitized
} }

97
src/automerge/CloudflareAdapter.ts
View File

@ -48,7 +48,15 @@ export class CloudflareAdapter {
// Focus on the store data which is what actually changes // Focus on the store data which is what actually changes
const storeData = doc.store || {} const storeData = doc.store || {}
const storeKeys = Object.keys(storeData).sort() const storeKeys = Object.keys(storeData).sort()
const storeString = JSON.stringify(storeData, storeKeys)
// CRITICAL FIX: JSON.stringify's second parameter when it's an array is a replacer
// that only includes those properties. We need to stringify the entire store object.
// To ensure stable ordering, create a new object with sorted keys
const sortedStore: any = {}
for (const key of storeKeys) {
sortedStore[key] = storeData[key]
}
const storeString = JSON.stringify(sortedStore)
// Simple hash function (you could use a more sophisticated one if needed) // Simple hash function (you could use a more sophisticated one if needed)
let hash = 0 let hash = 0
@ -158,6 +166,7 @@ export class CloudflareNetworkAdapter extends NetworkAdapter {
private websocket: WebSocket | null = null private websocket: WebSocket | null = null
private roomId: string | null = null private roomId: string | null = null
public peerId: PeerId | undefined = undefined public peerId: PeerId | undefined = undefined
public sessionId: string | null = null // Track our session ID
private readyPromise: Promise<void> private readyPromise: Promise<void>
private readyResolve: (() => void) | null = null private readyResolve: (() => void) | null = null
private keepAliveInterval: NodeJS.Timeout | null = null private keepAliveInterval: NodeJS.Timeout | null = null
@ -167,12 +176,19 @@ export class CloudflareNetworkAdapter extends NetworkAdapter {
private reconnectDelay: number = 1000 private reconnectDelay: number = 1000
private isConnecting: boolean = false private isConnecting: boolean = false
private onJsonSyncData?: (data: any) => void private onJsonSyncData?: (data: any) => void
private onPresenceUpdate?: (userId: string, data: any, senderId?: string, userName?: string, userColor?: string) => void
constructor(workerUrl: string, roomId?: string, onJsonSyncData?: (data: any) => void) { constructor(
workerUrl: string,
roomId?: string,
onJsonSyncData?: (data: any) => void,
onPresenceUpdate?: (userId: string, data: any, senderId?: string, userName?: string, userColor?: string) => void
) {
super() super()
this.workerUrl = workerUrl this.workerUrl = workerUrl
this.roomId = roomId || 'default-room' this.roomId = roomId || 'default-room'
this.onJsonSyncData = onJsonSyncData this.onJsonSyncData = onJsonSyncData
this.onPresenceUpdate = onPresenceUpdate
this.readyPromise = new Promise((resolve) => { this.readyPromise = new Promise((resolve) => {
this.readyResolve = resolve this.readyResolve = resolve
}) })
@ -201,11 +217,13 @@ export class CloudflareNetworkAdapter extends NetworkAdapter {
// Use the room ID from constructor or default // Use the room ID from constructor or default
// Add sessionId as a query parameter as required by AutomergeDurableObject // Add sessionId as a query parameter as required by AutomergeDurableObject
const sessionId = peerId || `session-${Date.now()}-${Math.random().toString(36).substr(2, 9)}` const sessionId = peerId || `session-${Date.now()}-${Math.random().toString(36).substr(2, 9)}`
this.sessionId = sessionId // Store our session ID for filtering echoes
// Convert https:// to wss:// or http:// to ws:// // Convert https:// to wss:// or http:// to ws://
const protocol = this.workerUrl.startsWith('https://') ? 'wss://' : 'ws://' const protocol = this.workerUrl.startsWith('https://') ? 'wss://' : 'ws://'
const baseUrl = this.workerUrl.replace(/^https?:\/\//, '') const baseUrl = this.workerUrl.replace(/^https?:\/\//, '')
const wsUrl = `${protocol}${baseUrl}/connect/${this.roomId}?sessionId=${sessionId}` const wsUrl = `${protocol}${baseUrl}/connect/${this.roomId}?sessionId=${sessionId}`
this.isConnecting = true this.isConnecting = true
// Add a small delay to ensure the server is ready // Add a small delay to ensure the server is ready
@ -252,19 +270,32 @@ export class CloudflareNetworkAdapter extends NetworkAdapter {
} else { } else {
// Handle text messages (our custom protocol for backward compatibility) // Handle text messages (our custom protocol for backward compatibility)
const message = JSON.parse(event.data) const message = JSON.parse(event.data)
console.log('🔌 CloudflareAdapter: Received WebSocket message:', message.type)
// Only log non-presence messages to reduce console spam
if (message.type !== 'presence' && message.type !== 'pong') {
console.log('🔌 CloudflareAdapter: Received WebSocket message:', message.type)
}
// Handle ping/pong messages for keep-alive // Handle ping/pong messages for keep-alive
if (message.type === 'ping') { if (message.type === 'ping') {
this.sendPong() this.sendPong()
return return
} }
// Handle test messages // Handle test messages
if (message.type === 'test') { if (message.type === 'test') {
console.log('🔌 CloudflareAdapter: Received test message:', message.message) console.log('🔌 CloudflareAdapter: Received test message:', message.message)
return return
} }
// Handle presence updates from other clients
if (message.type === 'presence') {
// Pass senderId, userName, and userColor so we can create proper instance_presence records
if (this.onPresenceUpdate && message.userId && message.data) {
this.onPresenceUpdate(message.userId, message.data, message.senderId, message.userName, message.userColor)
}
return
}
// Convert the message to the format expected by Automerge // Convert the message to the format expected by Automerge
if (message.type === 'sync' && message.data) { if (message.type === 'sync' && message.data) {
@ -275,14 +306,20 @@ export class CloudflareNetworkAdapter extends NetworkAdapter {
documentIdType: typeof message.documentId documentIdType: typeof message.documentId
}) })
// JSON sync is deprecated - all data flows through Automerge sync protocol // JSON sync for real-time collaboration
// Old format content is converted server-side and saved to R2 in Automerge format // When we receive TLDraw changes from other clients, apply them locally
// Skip JSON sync messages - they should not be sent anymore
const isJsonDocumentData = message.data && typeof message.data === 'object' && message.data.store const isJsonDocumentData = message.data && typeof message.data === 'object' && message.data.store
if (isJsonDocumentData) { if (isJsonDocumentData) {
console.warn('⚠️ CloudflareAdapter: Received JSON sync message (deprecated). Ignoring - all data should flow through Automerge sync protocol.') console.log('📥 CloudflareAdapter: Received JSON sync message with store data')
return // Don't process JSON sync messages
// Call the JSON sync callback to apply changes
if (this.onJsonSyncData) {
this.onJsonSyncData(message.data)
} else {
console.warn('⚠️ No JSON sync callback registered')
}
return // JSON sync handled
} }
// Validate documentId - Automerge requires a valid Automerge URL format // Validate documentId - Automerge requires a valid Automerge URL format
@ -368,19 +405,42 @@ export class CloudflareNetworkAdapter extends NetworkAdapter {
} }
send(message: Message): void { send(message: Message): void {
// Only log non-presence messages to reduce console spam
if (message.type !== 'presence') {
console.log('📤 CloudflareAdapter.send() called:', {
messageType: message.type,
dataType: (message as any).data?.constructor?.name || typeof (message as any).data,
dataLength: (message as any).data?.byteLength || (message as any).data?.length,
documentId: (message as any).documentId,
hasTargetId: !!message.targetId,
hasSenderId: !!message.senderId
})
}
if (this.websocket && this.websocket.readyState === WebSocket.OPEN) { if (this.websocket && this.websocket.readyState === WebSocket.OPEN) {
// Check if this is a binary sync message from Automerge Repo // Check if this is a binary sync message from Automerge Repo
if (message.type === 'sync' && (message as any).data instanceof ArrayBuffer) { if (message.type === 'sync' && (message as any).data instanceof ArrayBuffer) {
console.log('🔌 CloudflareAdapter: Sending binary sync message (Automerge protocol)') console.log('📤 CloudflareAdapter: Sending binary sync message (Automerge protocol)', {
dataLength: (message as any).data.byteLength,
documentId: (message as any).documentId,
targetId: message.targetId
})
// Send binary data directly for Automerge's native sync protocol // Send binary data directly for Automerge's native sync protocol
this.websocket.send((message as any).data) this.websocket.send((message as any).data)
} else if (message.type === 'sync' && (message as any).data instanceof Uint8Array) { } else if (message.type === 'sync' && (message as any).data instanceof Uint8Array) {
console.log('🔌 CloudflareAdapter: Sending Uint8Array sync message (Automerge protocol)') console.log('📤 CloudflareAdapter: Sending Uint8Array sync message (Automerge protocol)', {
dataLength: (message as any).data.length,
documentId: (message as any).documentId,
targetId: message.targetId
})
// Convert Uint8Array to ArrayBuffer and send // Convert Uint8Array to ArrayBuffer and send
this.websocket.send((message as any).data.buffer) this.websocket.send((message as any).data.buffer)
} else { } else {
// Handle text-based messages (backward compatibility and control messages) // Handle text-based messages (backward compatibility and control messages)
console.log('Sending WebSocket message:', message.type) // Only log non-presence messages
if (message.type !== 'presence') {
console.log('📤 Sending WebSocket message:', message.type)
}
// Debug: Log patch content if it's a patch message // Debug: Log patch content if it's a patch message
if (message.type === 'patch' && (message as any).patches) { if (message.type === 'patch' && (message as any).patches) {
console.log('🔍 Sending patches:', (message as any).patches.length, 'patches') console.log('🔍 Sending patches:', (message as any).patches.length, 'patches')
@ -394,6 +454,13 @@ export class CloudflareNetworkAdapter extends NetworkAdapter {
} }
this.websocket.send(JSON.stringify(message)) this.websocket.send(JSON.stringify(message))
} }
} else {
if (message.type !== 'presence') {
console.warn('⚠️ CloudflareAdapter: Cannot send message - WebSocket not open', {
messageType: message.type,
readyState: this.websocket?.readyState
})
}
} }
} }

33
src/automerge/MinimalSanitization.ts
View File

@ -20,7 +20,38 @@ function minimalSanitizeRecord(record: any): any {
if (typeof sanitized.isLocked !== 'boolean') sanitized.isLocked = false if (typeof sanitized.isLocked !== 'boolean') sanitized.isLocked = false
if (typeof sanitized.opacity !== 'number') sanitized.opacity = 1 if (typeof sanitized.opacity !== 'number') sanitized.opacity = 1
if (!sanitized.meta || typeof sanitized.meta !== 'object') sanitized.meta = {} if (!sanitized.meta || typeof sanitized.meta !== 'object') sanitized.meta = {}
if (!sanitized.index) sanitized.index = 'a1' // NOTE: Index assignment is handled by assignSequentialIndices() during format conversion
// Here we validate using tldraw's fractional indexing rules
// The first letter encodes integer part length: a=1 digit, b=2 digits, c=3 digits, etc.
// Examples: "a0"-"a9", "b10"-"b99", "c100"-"c999", with optional fraction "a1V4rr"
// Invalid: "b1" (b expects 2 digits but has 1)
if (!sanitized.index || typeof sanitized.index !== 'string' || sanitized.index.length === 0) {
sanitized.index = 'a1'
} else {
// Validate fractional indexing format
let isValid = false
const prefix = sanitized.index[0]
const rest = sanitized.index.slice(1)
if (/^[a-zA-Z]/.test(sanitized.index) && /^[a-zA-Z][a-zA-Z0-9]+$/.test(sanitized.index)) {
if (prefix >= 'a' && prefix <= 'z') {
// Calculate expected minimum digit count: a=1, b=2, c=3, etc.
const expectedDigits = prefix.charCodeAt(0) - 'a'.charCodeAt(0) + 1
const integerMatch = rest.match(/^(\d+)/)
if (integerMatch && integerMatch[1].length >= expectedDigits) {
isValid = true
}
} else if (prefix >= 'A' && prefix <= 'Z') {
// Uppercase for negative/special indices - allow
isValid = true
}
}
if (!isValid) {
console.warn(`⚠️ MinimalSanitization: Invalid index format "${sanitized.index}" for shape ${sanitized.id}`)
sanitized.index = 'a1'
}
}
if (!sanitized.parentId) sanitized.parentId = 'page:page' if (!sanitized.parentId) sanitized.parentId = 'page:page'
// Ensure props object exists // Ensure props object exists

4
src/automerge/README.md
View File

@ -47,6 +47,6 @@ To switch from TLdraw sync to Automerge sync:
1. Update the Board component to use `useAutomergeSync` 1. Update the Board component to use `useAutomergeSync`
2. Deploy the new worker with Automerge Durable Object 2. Deploy the new worker with Automerge Durable Object
3. Update the URI to use `/automerge/connect/` instead of `/connect/` 3. The CloudflareAdapter will automatically connect to `/connect/{roomId}` via WebSocket
The migration is backward compatible - existing TLdraw sync will continue to work while you test the new system. The migration is backward compatible - the system will handle both legacy and new document formats.

107
src/automerge/TLStoreToAutomerge.ts
View File

@ -144,19 +144,97 @@ function sanitizeRecord(record: TLRecord): TLRecord {
console.warn(`🔧 TLStoreToAutomerge: Error checking richText for shape ${sanitized.id}:`, e) console.warn(`🔧 TLStoreToAutomerge: Error checking richText for shape ${sanitized.id}:`, e)
} }
// CRITICAL: Extract arrow text BEFORE deep copy to handle RichText instances properly
// Arrow text should be a string, but might be a RichText object in edge cases
let arrowTextValue: any = undefined
if (sanitized.type === 'arrow') {
try {
const props = sanitized.props || {}
if ('text' in props) {
try {
// Use Object.getOwnPropertyDescriptor to safely check if it's a getter
const descriptor = Object.getOwnPropertyDescriptor(props, 'text')
let textValue: any = undefined
if (descriptor && descriptor.get) {
// It's a getter - try to call it safely
try {
textValue = descriptor.get.call(props)
} catch (getterError) {
console.warn(`🔧 TLStoreToAutomerge: Error calling text getter for arrow ${sanitized.id}:`, getterError)
textValue = undefined
}
} else {
// It's a regular property - access it directly
textValue = (props as any).text
}
// Now process the value
if (textValue !== undefined && textValue !== null) {
// If it's a string, use it directly
if (typeof textValue === 'string') {
arrowTextValue = textValue
}
// If it's a RichText object, extract the text content
else if (typeof textValue === 'object' && textValue !== null) {
// Try to extract text from RichText object
try {
const serialized = JSON.parse(JSON.stringify(textValue))
// If it has content array, extract text from it
if (Array.isArray(serialized.content)) {
// Extract text from RichText content
const extractText = (content: any[]): string => {
return content.map((item: any) => {
if (item.type === 'text' && item.text) {
return item.text
} else if (item.content && Array.isArray(item.content)) {
return extractText(item.content)
}
return ''
}).join('')
}
arrowTextValue = extractText(serialized.content)
} else {
// Fallback: try to get text property
arrowTextValue = serialized.text || ''
}
} catch (serializeError) {
// If serialization fails, try to extract manually
if ((textValue as any).text && typeof (textValue as any).text === 'string') {
arrowTextValue = (textValue as any).text
} else {
arrowTextValue = String(textValue)
}
}
}
// For other types, convert to string
else {
arrowTextValue = String(textValue)
}
}
} catch (e) {
console.warn(`🔧 TLStoreToAutomerge: Error extracting text for arrow ${sanitized.id}:`, e)
arrowTextValue = undefined
}
}
} catch (e) {
console.warn(`🔧 TLStoreToAutomerge: Error checking text for arrow ${sanitized.id}:`, e)
}
}
// CRITICAL: For all shapes, ensure props is a deep mutable copy to preserve all properties // CRITICAL: For all shapes, ensure props is a deep mutable copy to preserve all properties
// This is essential for custom shapes like ObsNote and for preserving richText in geo shapes // This is essential for custom shapes like ObsNote and for preserving richText in geo shapes
// Use JSON parse/stringify to create a deep copy of nested objects (like richText.content) // Use JSON parse/stringify to create a deep copy of nested objects (like richText.content)
// Remove richText temporarily to avoid serialization issues // Remove richText and arrow text temporarily to avoid serialization issues
try { try {
const propsWithoutRichText: any = {} const propsWithoutSpecial: any = {}
// Copy all props except richText // Copy all props except richText and arrow text (if extracted)
for (const key in sanitized.props) { for (const key in sanitized.props) {
if (key !== 'richText') { if (key !== 'richText' && !(sanitized.type === 'arrow' && key === 'text' && arrowTextValue !== undefined)) {
propsWithoutRichText[key] = (sanitized.props as any)[key] propsWithoutSpecial[key] = (sanitized.props as any)[key]
} }
} }
sanitized.props = JSON.parse(JSON.stringify(propsWithoutRichText)) sanitized.props = JSON.parse(JSON.stringify(propsWithoutSpecial))
} catch (e) { } catch (e) {
console.warn(`🔧 TLStoreToAutomerge: Error deep copying props for shape ${sanitized.id}:`, e) console.warn(`🔧 TLStoreToAutomerge: Error deep copying props for shape ${sanitized.id}:`, e)
// Fallback: just copy props without deep copy // Fallback: just copy props without deep copy
@ -164,6 +242,9 @@ function sanitizeRecord(record: TLRecord): TLRecord {
if (richTextValue !== undefined) { if (richTextValue !== undefined) {
delete (sanitized.props as any).richText delete (sanitized.props as any).richText
} }
if (arrowTextValue !== undefined) {
delete (sanitized.props as any).text
}
} }
// CRITICAL: For geo shapes, move w/h/geo from top-level to props (required by TLDraw schema) // CRITICAL: For geo shapes, move w/h/geo from top-level to props (required by TLDraw schema)
@ -210,11 +291,17 @@ function sanitizeRecord(record: TLRecord): TLRecord {
// CRITICAL: For arrow shapes, preserve text property // CRITICAL: For arrow shapes, preserve text property
if (sanitized.type === 'arrow') { if (sanitized.type === 'arrow') {
// CRITICAL: Preserve text property - only set default if truly missing (preserve empty strings and all other values) // CRITICAL: Restore extracted text value if available, otherwise preserve existing text
if ((sanitized.props as any).text === undefined || (sanitized.props as any).text === null) { if (arrowTextValue !== undefined) {
(sanitized.props as any).text = '' // Use the extracted text value (handles RichText objects by extracting text content)
(sanitized.props as any).text = arrowTextValue
} else {
// CRITICAL: Preserve text property - only set default if truly missing (preserve empty strings and all other values)
if ((sanitized.props as any).text === undefined || (sanitized.props as any).text === null) {
(sanitized.props as any).text = ''
}
// Note: We preserve text even if it's an empty string - that's a valid value
} }
// Note: We preserve text even if it's an empty string - that's a valid value
} }
// CRITICAL: For note shapes, preserve richText property (required for note shapes) // CRITICAL: For note shapes, preserve richText property (required for note shapes)

6
src/automerge/documentIdMapping.ts
View File

@ -115,7 +115,7 @@ export async function saveDocumentId(roomId: string, documentId: string): Promis
} }
request.onsuccess = () => { request.onsuccess = () => {
console.log(`📝 Saved document mapping: ${roomId} → ${documentId}`) console.log(`Saved document mapping: ${roomId} -> ${documentId}`)
resolve() resolve()
} }
}) })
@ -171,7 +171,7 @@ export async function deleteDocumentMapping(roomId: string): Promise<void> {
} }
request.onsuccess = () => { request.onsuccess = () => {
console.log(`🗑️ Deleted document mapping for: ${roomId}`) console.log(`Deleted document mapping for: ${roomId}`)
resolve() resolve()
} }
}) })
@ -238,7 +238,7 @@ export async function cleanupOldMappings(maxAgeDays: number = 30): Promise<numbe
deletedCount++ deletedCount++
cursor.continue() cursor.continue()
} else { } else {
console.log(`🧹 Cleaned up ${deletedCount} old document mappings`) console.log(`Cleaned up ${deletedCount} old document mappings`)
resolve(deletedCount) resolve(deletedCount)
} }
} }

1377
src/automerge/useAutomergeStoreV2.ts
View File

File diff suppressed because it is too large Load Diff

951
src/automerge/useAutomergeSyncRepo.ts
View File

File diff suppressed because it is too large Load Diff

450
src/components/FathomMeetingsPanel.tsx
View File

@ -1,45 +1,67 @@
import React, { useState, useEffect } from 'react' import React, { useState, useEffect, useContext, useRef } from 'react'
import { useEditor } from 'tldraw' import { useEditor } from 'tldraw'
import { createShapeId } from 'tldraw' import { createShapeId } from 'tldraw'
import { WORKER_URL, LOCAL_WORKER_URL } from '../constants/workerUrl' import { WORKER_URL, LOCAL_WORKER_URL } from '../constants/workerUrl'
import { getFathomApiKey, saveFathomApiKey, removeFathomApiKey } from '../lib/fathomApiKey'
import { AuthContext } from '../context/AuthContext'
interface FathomMeeting { interface FathomMeeting {
id: string recording_id: number
title: string title: string
meeting_title?: string
url: string url: string
share_url?: string
created_at: string created_at: string
duration: number scheduled_start_time?: string
summary?: { scheduled_end_time?: string
markdown_formatted: string recording_start_time?: string
recording_end_time?: string
transcript?: any[]
transcript_language?: string
default_summary?: {
template_name?: string
markdown_formatted?: string
} }
action_items?: any[]
calendar_invitees?: Array<{
name: string
email: string
is_external: boolean
}>
recorded_by?: {
name: string
email: string
team?: string
}
call_id?: string | number
id?: string | number
} }
interface FathomMeetingsPanelProps { interface FathomMeetingsPanelProps {
onClose: () => void onClose?: () => void
onMeetingSelect?: (meeting: FathomMeeting, options: { summary: boolean; transcript: boolean; actionItems: boolean; video: boolean }, format: 'fathom' | 'note') => void
shapeMode?: boolean shapeMode?: boolean
} }
export function FathomMeetingsPanel({ onClose, shapeMode = false }: FathomMeetingsPanelProps) { export function FathomMeetingsPanel({ onClose, onMeetingSelect, shapeMode = false }: FathomMeetingsPanelProps) {
const editor = useEditor() const editor = useEditor()
// Safely get auth context - may not be available during SVG export
const authContext = useContext(AuthContext)
const fallbackSession = {
username: undefined as string | undefined,
}
const session = authContext?.session || fallbackSession
const [apiKey, setApiKey] = useState('') const [apiKey, setApiKey] = useState('')
const [showApiKeyInput, setShowApiKeyInput] = useState(false) const [showApiKeyInput, setShowApiKeyInput] = useState(false)
const [meetings, setMeetings] = useState<FathomMeeting[]>([]) const [meetings, setMeetings] = useState<FathomMeeting[]>([])
const [loading, setLoading] = useState(false) const [loading, setLoading] = useState(false)
const [error, setError] = useState<string | null>(null) const [error, setError] = useState<string | null>(null)
// Removed dropdown state - using buttons instead
useEffect(() => { const fetchMeetings = async (keyToUse?: string) => {
// Check if API key is already stored const key = keyToUse || apiKey
const storedApiKey = localStorage.getItem('fathom_api_key') if (!key) {
if (storedApiKey) {
setApiKey(storedApiKey)
fetchMeetings()
} else {
setShowApiKeyInput(true)
}
}, [])
const fetchMeetings = async () => {
if (!apiKey) {
setError('Please enter your Fathom API key') setError('Please enter your Fathom API key')
return return
} }
@ -53,7 +75,7 @@ export function FathomMeetingsPanel({ onClose, shapeMode = false }: FathomMeetin
try { try {
response = await fetch(`${WORKER_URL}/fathom/meetings`, { response = await fetch(`${WORKER_URL}/fathom/meetings`, {
headers: { headers: {
'Authorization': `Bearer ${apiKey}`, 'X-Api-Key': key,
'Content-Type': 'application/json' 'Content-Type': 'application/json'
} }
}) })
@ -61,7 +83,7 @@ export function FathomMeetingsPanel({ onClose, shapeMode = false }: FathomMeetin
console.log('Production worker failed, trying local worker...') console.log('Production worker failed, trying local worker...')
response = await fetch(`${LOCAL_WORKER_URL}/fathom/meetings`, { response = await fetch(`${LOCAL_WORKER_URL}/fathom/meetings`, {
headers: { headers: {
'Authorization': `Bearer ${apiKey}`, 'X-Api-Key': key,
'Content-Type': 'application/json' 'Content-Type': 'application/json'
} }
}) })
@ -91,28 +113,169 @@ export function FathomMeetingsPanel({ onClose, shapeMode = false }: FathomMeetin
const saveApiKey = () => { const saveApiKey = () => {
if (apiKey) { if (apiKey) {
localStorage.setItem('fathom_api_key', apiKey) saveFathomApiKey(apiKey, session.username)
setShowApiKeyInput(false) setShowApiKeyInput(false)
fetchMeetings() fetchMeetings(apiKey)
} }
} }
const addMeetingToCanvas = async (meeting: FathomMeeting) => { // Track if we've already loaded meetings for the current user to prevent multiple API calls
const hasLoadedRef = useRef<string | undefined>(undefined)
const hasMountedRef = useRef(false)
useEffect(() => {
// Only run once on mount, don't re-fetch when session.username changes
if (hasMountedRef.current) {
return // Already loaded, don't refresh
}
hasMountedRef.current = true
// Always check user profile first for API key, then fallback to global storage
const username = session.username
const storedApiKey = getFathomApiKey(username)
if (storedApiKey) {
setApiKey(storedApiKey)
setShowApiKeyInput(false)
// Automatically fetch meetings when API key is available
// Only fetch once per user to prevent unnecessary API calls
if (hasLoadedRef.current !== username) {
hasLoadedRef.current = username
fetchMeetings(storedApiKey)
}
} else {
setShowApiKeyInput(true)
hasLoadedRef.current = undefined
}
}, []) // Empty dependency array - only run once on mount
// Handler for individual data type buttons - creates shapes directly
const handleDataButtonClick = async (meeting: FathomMeeting, dataType: 'summary' | 'transcript' | 'actionItems' | 'video') => {
// Log to verify the correct meeting is being used
console.log('🔵 handleDataButtonClick called with meeting:', {
recording_id: meeting.recording_id,
title: meeting.title,
dataType
})
if (!onMeetingSelect) {
// Fallback for non-browser mode
const options = {
summary: dataType === 'summary',
transcript: dataType === 'transcript',
actionItems: dataType === 'actionItems',
video: dataType === 'video',
}
await addMeetingToCanvas(meeting, options)
return
}
// Browser mode - use callback with specific data type
// IMPORTANT: Pass the meeting object directly to ensure each button uses its own meeting's data
const options = {
summary: dataType === 'summary',
transcript: dataType === 'transcript',
actionItems: dataType === 'actionItems',
video: dataType === 'video',
}
// Always use 'note' format for summary, transcript, and action items (same behavior)
// Video opens URL directly, so format doesn't matter for it
const format = 'note'
onMeetingSelect(meeting, options, format)
}
const formatMeetingDataAsMarkdown = (fullMeeting: any, meeting: FathomMeeting, options: { summary: boolean; transcript: boolean; actionItems: boolean; video: boolean }): string => {
const parts: string[] = []
// Title
parts.push(`# ${fullMeeting.title || meeting.meeting_title || meeting.title || 'Meeting'}\n`)
// Video link if selected
if (options.video && (fullMeeting.url || meeting.url)) {
parts.push(`**Video:** [Watch Recording](${fullMeeting.url || meeting.url})\n`)
}
// Summary if selected
if (options.summary && fullMeeting.default_summary?.markdown_formatted) {
parts.push(`## Summary\n\n${fullMeeting.default_summary.markdown_formatted}\n`)
}
// Action Items if selected
if (options.actionItems && fullMeeting.action_items && fullMeeting.action_items.length > 0) {
parts.push(`## Action Items\n\n`)
fullMeeting.action_items.forEach((item: any) => {
const description = item.description || item.text || ''
const assignee = item.assignee?.name || item.assignee || ''
const dueDate = item.due_date || ''
parts.push(`- [ ] ${description}`)
if (assignee) parts[parts.length - 1] += ` (@${assignee})`
if (dueDate) parts[parts.length - 1] += ` - Due: ${dueDate}`
parts[parts.length - 1] += '\n'
})
parts.push('\n')
}
// Transcript if selected
if (options.transcript && fullMeeting.transcript && fullMeeting.transcript.length > 0) {
parts.push(`## Transcript\n\n`)
fullMeeting.transcript.forEach((entry: any) => {
const speaker = entry.speaker?.display_name || 'Unknown'
const text = entry.text || ''
const timestamp = entry.timestamp || ''
if (timestamp) {
parts.push(`**${speaker}** (${timestamp}): ${text}\n\n`)
} else {
parts.push(`**${speaker}**: ${text}\n\n`)
}
})
}
return parts.join('')
}
const addMeetingToCanvas = async (meeting: FathomMeeting, options: { summary: boolean; transcript: boolean; actionItems: boolean; video: boolean }) => {
try { try {
// If video is selected, just open the Fathom URL directly
if (options.video) {
// Try multiple sources for the correct video URL
// The Fathom API may provide url, share_url, or we may need to construct from call_id or id
const callId = meeting.call_id ||
meeting.id ||
meeting.recording_id
// Check if URL fields contain valid meeting URLs (contain /calls/)
const isValidMeetingUrl = (url: string) => url && url.includes('/calls/')
// Prioritize valid meeting URLs, then construct from call ID
const videoUrl = (meeting.url && isValidMeetingUrl(meeting.url)) ? meeting.url :
(meeting.share_url && isValidMeetingUrl(meeting.share_url)) ? meeting.share_url :
(callId ? `https://fathom.video/calls/${callId}` : null)
if (videoUrl) {
console.log('Opening Fathom video URL:', videoUrl, 'for meeting:', { callId, recording_id: meeting.recording_id })
window.open(videoUrl, '_blank', 'noopener,noreferrer')
} else {
console.error('Could not determine Fathom video URL for meeting:', meeting)
}
return
}
// Only fetch transcript if transcript is selected
const includeTranscript = options.transcript
// Fetch full meeting details // Fetch full meeting details
let response let response
try { try {
response = await fetch(`${WORKER_URL}/fathom/meetings/${meeting.id}`, { response = await fetch(`${WORKER_URL}/fathom/meetings/${meeting.recording_id}${includeTranscript ? '?include_transcript=true' : ''}`, {
headers: { headers: {
'Authorization': `Bearer ${apiKey}`, 'X-Api-Key': apiKey,
'Content-Type': 'application/json' 'Content-Type': 'application/json'
} }
}) })
} catch (error) { } catch (error) {
console.log('Production worker failed, trying local worker...') console.log('Production worker failed, trying local worker...')
response = await fetch(`${LOCAL_WORKER_URL}/fathom/meetings/${meeting.id}`, { response = await fetch(`${LOCAL_WORKER_URL}/fathom/meetings/${meeting.recording_id}${includeTranscript ? '?include_transcript=true' : ''}`, {
headers: { headers: {
'Authorization': `Bearer ${apiKey}`, 'X-Api-Key': apiKey,
'Content-Type': 'application/json' 'Content-Type': 'application/json'
} }
}) })
@ -125,41 +288,60 @@ export function FathomMeetingsPanel({ onClose, shapeMode = false }: FathomMeetin
const fullMeeting = await response.json() as any const fullMeeting = await response.json() as any
// Create Fathom transcript shape // If onMeetingSelect callback is provided, use it (browser mode - creates separate shapes)
if (onMeetingSelect) {
// Default to 'note' format for text data
onMeetingSelect(meeting, options, 'note')
// Browser stays open, don't close
return
}
// Fallback: create shape directly (for non-browser mode, like modal)
// Default to note format
const markdownContent = formatMeetingDataAsMarkdown(fullMeeting, meeting, options)
const title = fullMeeting.title || meeting.meeting_title || meeting.title || 'Fathom Meeting'
const shapeId = createShapeId() const shapeId = createShapeId()
editor.createShape({ editor.createShape({
id: shapeId, id: shapeId,
type: 'FathomTranscript', type: 'ObsNote',
x: 100, x: 100,
y: 100, y: 100,
props: { props: {
meetingId: fullMeeting.id || '', w: 400,
meetingTitle: fullMeeting.title || '', h: 500,
meetingUrl: fullMeeting.url || '', color: 'black',
summary: fullMeeting.default_summary?.markdown_formatted || '', size: 'm',
transcript: fullMeeting.transcript?.map((entry: any) => ({ font: 'sans',
speaker: entry.speaker?.display_name || 'Unknown', textAlign: 'start',
text: entry.text, scale: 1,
timestamp: entry.timestamp noteId: `fathom-${meeting.recording_id}`,
})) || [], title: title,
actionItems: fullMeeting.action_items?.map((item: any) => ({ content: markdownContent,
text: item.text, tags: ['fathom', 'meeting'],
assignee: item.assignee, showPreview: true,
dueDate: item.due_date backgroundColor: '#ffffff',
})) || [], textColor: '#000000',
isExpanded: false, isEditing: false,
showTranscript: true, editingContent: '',
showActionItems: true, isModified: false,
originalContent: markdownContent,
pinnedToView: false,
} }
}) })
onClose() // Only close if not in shape mode (browser stays open)
if (!shapeMode && onClose) {
onClose()
}
} catch (error) { } catch (error) {
console.error('Error adding meeting to canvas:', error) console.error('Error adding meeting to canvas:', error)
setError(`Failed to add meeting: ${(error as Error).message}`) setError(`Failed to add meeting: ${(error as Error).message}`)
} }
} }
// Removed dropdown click-outside handler - no longer needed with button-based interface
const formatDate = (dateString: string) => { const formatDate = (dateString: string) => {
return new Date(dateString).toLocaleDateString() return new Date(dateString).toLocaleDateString()
} }
@ -196,38 +378,22 @@ export function FathomMeetingsPanel({ onClose, shapeMode = false }: FathomMeetin
} }
const content = ( const content = (
<div style={contentStyle} onClick={(e) => shapeMode ? undefined : e.stopPropagation()}> <div
<div style={{ style={contentStyle}
display: 'flex', onClick={(e) => {
justifyContent: 'space-between', // Prevent clicks from interfering with shape selection or resetting data
alignItems: 'center', if (!shapeMode) {
marginBottom: '20px', e.stopPropagation()
paddingBottom: '10px', }
borderBottom: '1px solid #eee' // In shape mode, allow normal interaction but don't reset data
}}> }}
<h2 style={{ margin: 0, fontSize: '18px', fontWeight: 'bold' }}> onMouseDown={(e) => {
🎥 Fathom Meetings // Prevent shape deselection when clicking inside the browser content
</h2> if (shapeMode) {
<button e.stopPropagation()
onClick={(e) => { }
e.stopPropagation() }}
onClose() >
}}
style={{
background: 'none',
border: 'none',
fontSize: '20px',
cursor: 'pointer',
padding: '5px',
position: 'relative',
zIndex: 10002,
pointerEvents: 'auto'
}}
>
✕
</button>
</div>
{showApiKeyInput ? ( {showApiKeyInput ? (
<div> <div>
<p style={{ <p style={{
@ -269,7 +435,8 @@ export function FathomMeetingsPanel({ onClose, shapeMode = false }: FathomMeetin
cursor: apiKey ? 'pointer' : 'not-allowed', cursor: apiKey ? 'pointer' : 'not-allowed',
position: 'relative', position: 'relative',
zIndex: 10002, zIndex: 10002,
pointerEvents: 'auto' pointerEvents: 'auto',
touchAction: 'manipulation'
}} }}
> >
Save & Load Meetings Save & Load Meetings
@ -296,7 +463,7 @@ export function FathomMeetingsPanel({ onClose, shapeMode = false }: FathomMeetin
<> <>
<div style={{ display: 'flex', gap: '10px', marginBottom: '20px' }}> <div style={{ display: 'flex', gap: '10px', marginBottom: '20px' }}>
<button <button
onClick={fetchMeetings} onClick={() => fetchMeetings(apiKey)}
disabled={loading} disabled={loading}
style={{ style={{
padding: '8px 16px', padding: '8px 16px',
@ -314,9 +481,12 @@ export function FathomMeetingsPanel({ onClose, shapeMode = false }: FathomMeetin
</button> </button>
<button <button
onClick={() => { onClick={() => {
localStorage.removeItem('fathom_api_key') // Remove API key from user-specific storage
removeFathomApiKey(session.username)
setApiKey('') setApiKey('')
setMeetings([])
setShowApiKeyInput(true) setShowApiKeyInput(true)
hasLoadedRef.current = undefined
}} }}
style={{ style={{
padding: '8px 16px', padding: '8px 16px',
@ -363,7 +533,7 @@ export function FathomMeetingsPanel({ onClose, shapeMode = false }: FathomMeetin
) : ( ) : (
meetings.map((meeting) => ( meetings.map((meeting) => (
<div <div
key={meeting.id} key={meeting.recording_id}
style={{ style={{
border: '1px solid #e0e0e0', border: '1px solid #e0e0e0',
borderRadius: '6px', borderRadius: '6px',
@ -393,9 +563,11 @@ export function FathomMeetingsPanel({ onClose, shapeMode = false }: FathomMeetin
cursor: 'text' cursor: 'text'
}}> }}>
<div>📅 {formatDate(meeting.created_at)}</div> <div>📅 {formatDate(meeting.created_at)}</div>
<div>⏱️ Duration: {formatDuration(meeting.duration)}</div> <div>⏱️ Duration: {meeting.recording_start_time && meeting.recording_end_time
? formatDuration(Math.floor((new Date(meeting.recording_end_time).getTime() - new Date(meeting.recording_start_time).getTime()) / 1000))
: 'N/A'}</div>
</div> </div>
{meeting.summary && ( {meeting.default_summary?.markdown_formatted && (
<div style={{ <div style={{
fontSize: '11px', fontSize: '11px',
color: '#333', color: '#333',
@ -403,28 +575,91 @@ export function FathomMeetingsPanel({ onClose, shapeMode = false }: FathomMeetin
userSelect: 'text', userSelect: 'text',
cursor: 'text' cursor: 'text'
}}> }}>
<strong>Summary:</strong> {meeting.summary.markdown_formatted.substring(0, 100)}... <strong>Summary:</strong> {meeting.default_summary.markdown_formatted.substring(0, 100)}...
</div> </div>
)} )}
</div> </div>
<button <div style={{
onClick={() => addMeetingToCanvas(meeting)} display: 'flex',
style={{ flexDirection: 'row',
padding: '6px 12px', gap: '6px',
backgroundColor: '#28a745', marginLeft: '10px',
color: 'white', alignItems: 'center',
border: 'none', flexWrap: 'wrap'
borderRadius: '4px', }}>
cursor: 'pointer', <button
fontSize: '12px', onClick={() => handleDataButtonClick(meeting, 'summary')}
marginLeft: '10px', disabled={loading}
position: 'relative', style={{
zIndex: 10002, padding: '6px 12px',
pointerEvents: 'auto' backgroundColor: '#3b82f6',
}} color: 'white',
> border: 'none',
Add to Canvas borderRadius: '4px',
</button> cursor: loading ? 'not-allowed' : 'pointer',
fontSize: '11px',
whiteSpace: 'nowrap',
opacity: loading ? 0.6 : 1
}}
title="Add Summary as Note"
>
📄 Summary
</button>
<button
onClick={() => handleDataButtonClick(meeting, 'transcript')}
disabled={loading}
style={{
padding: '6px 12px',
backgroundColor: '#2563eb',
color: 'white',
border: 'none',
borderRadius: '4px',
cursor: loading ? 'not-allowed' : 'pointer',
fontSize: '11px',
whiteSpace: 'nowrap',
opacity: loading ? 0.6 : 1
}}
title="Add Transcript as Note"
>
📝 Transcript
</button>
<button
onClick={() => handleDataButtonClick(meeting, 'actionItems')}
disabled={loading}
style={{
padding: '6px 12px',
backgroundColor: '#1d4ed8',
color: 'white',
border: 'none',
borderRadius: '4px',
cursor: loading ? 'not-allowed' : 'pointer',
fontSize: '11px',
whiteSpace: 'nowrap',
opacity: loading ? 0.6 : 1
}}
title="Add Action Items as Note"
>
✅ Actions
</button>
<button
onClick={() => handleDataButtonClick(meeting, 'video')}
disabled={loading}
style={{
padding: '6px 12px',
backgroundColor: '#1e40af',
color: 'white',
border: 'none',
borderRadius: '4px',
cursor: loading ? 'not-allowed' : 'pointer',
fontSize: '11px',
whiteSpace: 'nowrap',
opacity: loading ? 0.6 : 1
}}
title="Add Video as Embed"
>
🎥 Video
</button>
</div>
</div> </div>
</div> </div>
)) ))
@ -477,3 +712,4 @@ export function FathomMeetingsPanel({ onClose, shapeMode = false }: FathomMeetin

74
src/components/HolonBrowser.tsx
View File

@ -49,14 +49,33 @@ export function HolonBrowser({ isOpen, onClose, onSelectHolon, shapeMode = false
setHolonInfo(null) setHolonInfo(null)
try { try {
// Validate that the holonId is a valid H3 index // Check if it's a valid H3 cell ID
if (!h3.isValidCell(holonId)) { const isH3Cell = h3.isValidCell(holonId)
throw new Error('Invalid H3 cell ID')
// Check if it's a numeric Holon ID (workspace/group identifier)
const isNumericId = /^\d{6,20}$/.test(holonId)
// Check if it's an alphanumeric identifier
const isAlphanumericId = /^[a-zA-Z0-9_-]{3,50}$/.test(holonId)
if (!isH3Cell && !isNumericId && !isAlphanumericId) {
throw new Error('Invalid Holon ID. Enter an H3 cell ID (e.g., 872a1070bffffff) or a numeric Holon ID (e.g., 1002848305066)')
} }
// Get holon information // Get holon information based on ID type
const resolution = h3.getResolution(holonId) let resolution: number
const [lat, lng] = h3.cellToLatLng(holonId) let lat: number
let lng: number
if (isH3Cell) {
resolution = h3.getResolution(holonId)
;[lat, lng] = h3.cellToLatLng(holonId)
} else {
// For non-H3 IDs, use default values
resolution = -1 // Indicates non-geospatial holon
lat = 0
lng = 0
}
// Try to get metadata from the holon // Try to get metadata from the holon
let metadata = null let metadata = null
@ -101,7 +120,9 @@ export function HolonBrowser({ isOpen, onClose, onSelectHolon, shapeMode = false
latitude: lat, latitude: lat,
longitude: lng, longitude: lng,
resolution: resolution, resolution: resolution,
resolutionName: HoloSphereService.getResolutionName(resolution), resolutionName: resolution >= 0
? HoloSphereService.getResolutionName(resolution)
: 'Workspace / Group',
data: {}, data: {},
lastUpdated: metadata?.lastUpdated || Date.now() lastUpdated: metadata?.lastUpdated || Date.now()
} }
@ -192,7 +213,7 @@ export function HolonBrowser({ isOpen, onClose, onSelectHolon, shapeMode = false
</button> </button>
</div> </div>
<p className="text-sm text-gray-600 mt-2"> <p className="text-sm text-gray-600 mt-2">
Enter a Holon ID to browse its data and import it to your canvas Enter a Holon ID (numeric like 1002848305066 or H3 cell like 872a1070bffffff) to browse its data
</p> </p>
</div> </div>
)} )}
@ -210,7 +231,7 @@ export function HolonBrowser({ isOpen, onClose, onSelectHolon, shapeMode = false
value={holonId} value={holonId}
onChange={(e) => setHolonId(e.target.value)} onChange={(e) => setHolonId(e.target.value)}
onKeyDown={handleKeyDown} onKeyDown={handleKeyDown}
placeholder="e.g., 1002848305066" placeholder="e.g., 1002848305066 or 872a1070bffffff"
className="flex-1 px-3 py-2 border border-gray-300 rounded-md focus:outline-none focus:ring-2 focus:ring-blue-500 z-[10001] relative" className="flex-1 px-3 py-2 border border-gray-300 rounded-md focus:outline-none focus:ring-2 focus:ring-blue-500 z-[10001] relative"
disabled={isLoading} disabled={isLoading}
style={{ zIndex: 10001 }} style={{ zIndex: 10001 }}
@ -237,18 +258,29 @@ export function HolonBrowser({ isOpen, onClose, onSelectHolon, shapeMode = false
</h3> </h3>
<div className="grid grid-cols-1 md:grid-cols-2 gap-4 mb-4"> <div className="grid grid-cols-1 md:grid-cols-2 gap-4 mb-4">
<div> {holonInfo.resolution >= 0 ? (
<p className="text-sm text-gray-600">Coordinates</p> <>
<p className="font-mono text-sm"> <div>
{holonInfo.latitude.toFixed(6)}, {holonInfo.longitude.toFixed(6)} <p className="text-sm text-gray-600">Coordinates</p>
</p> <p className="font-mono text-sm">
</div> {holonInfo.latitude.toFixed(6)}, {holonInfo.longitude.toFixed(6)}
<div> </p>
<p className="text-sm text-gray-600">Resolution</p> </div>
<p className="text-sm"> <div>
{holonInfo.resolutionName} (Level {holonInfo.resolution}) <p className="text-sm text-gray-600">Resolution</p>
</p> <p className="text-sm">
</div> {holonInfo.resolutionName} (Level {holonInfo.resolution})
</p>
</div>
</>
) : (
<div>
<p className="text-sm text-gray-600">Type</p>
<p className="text-sm font-medium text-green-600">
{holonInfo.resolutionName}
</p>
</div>
)}
<div> <div>
<p className="text-sm text-gray-600">Holon ID</p> <p className="text-sm text-gray-600">Holon ID</p>
<p className="font-mono text-xs break-all">{holonInfo.id}</p> <p className="font-mono text-xs break-all">{holonInfo.id}</p>

380
src/components/StandardizedToolWrapper.tsx
View File

@ -1,4 +1,29 @@
import React, { useState, ReactNode } from 'react' import React, { useState, ReactNode, useEffect, useRef, useMemo } from 'react'
// Hook to detect dark mode
function useIsDarkMode() {
const [isDark, setIsDark] = useState(() => {
if (typeof document !== 'undefined') {
return document.documentElement.classList.contains('dark')
}
return false
})
useEffect(() => {
const observer = new MutationObserver((mutations) => {
mutations.forEach((mutation) => {
if (mutation.attributeName === 'class') {
setIsDark(document.documentElement.classList.contains('dark'))
}
})
})
observer.observe(document.documentElement, { attributes: true })
return () => observer.disconnect()
}, [])
return isDark
}
export interface StandardizedToolWrapperProps { export interface StandardizedToolWrapperProps {
/** The title to display in the header */ /** The title to display in the header */
@ -25,6 +50,16 @@ export interface StandardizedToolWrapperProps {
editor?: any editor?: any
/** Shape ID for selection handling */ /** Shape ID for selection handling */
shapeId?: string shapeId?: string
/** Whether the shape is pinned to view */
isPinnedToView?: boolean
/** Callback when pin button is clicked */
onPinToggle?: () => void
/** Tags to display at the bottom of the shape */
tags?: string[]
/** Callback when tags are updated */
onTagsChange?: (tags: string[]) => void
/** Whether tags can be edited */
tagsEditable?: boolean
} }
/** /**
@ -44,9 +79,70 @@ export const StandardizedToolWrapper: React.FC<StandardizedToolWrapperProps> = (
headerContent, headerContent,
editor, editor,
shapeId, shapeId,
isPinnedToView = false,
onPinToggle,
tags = [],
onTagsChange,
tagsEditable = true,
}) => { }) => {
const [isHoveringHeader, setIsHoveringHeader] = useState(false) const [isHoveringHeader, setIsHoveringHeader] = useState(false)
const [isEditingTags, setIsEditingTags] = useState(false)
const [editingTagInput, setEditingTagInput] = useState('')
const tagInputRef = useRef<HTMLInputElement>(null)
const isDarkMode = useIsDarkMode()
// Dark mode aware colors
const colors = useMemo(() => isDarkMode ? {
contentBg: '#1a1a1a',
tagsBg: '#252525',
tagsBorder: '#404040',
tagBg: '#4a5568',
tagText: '#e4e4e4',
addTagBg: '#4a5568',
inputBg: '#333333',
inputBorder: '#555555',
} : {
contentBg: 'white',
tagsBg: '#f8f9fa',
tagsBorder: '#e0e0e0',
tagBg: '#6b7280',
tagText: 'white',
addTagBg: '#9ca3af',
inputBg: 'white',
inputBorder: '#9ca3af',
}, [isDarkMode])
// Bring selected shape to front when it becomes selected
useEffect(() => {
if (editor && shapeId && isSelected) {
try {
// Bring the shape to the front by updating its index
// Note: sendToFront doesn't exist in this version of tldraw
const allShapes = editor.getCurrentPageShapes()
let highestIndex = 'a0'
for (const s of allShapes) {
if (s.index && typeof s.index === 'string' && s.index > highestIndex) {
highestIndex = s.index
}
}
const shape = editor.getShape(shapeId)
if (shape) {
const match = highestIndex.match(/^([a-z])(\d+)$/)
if (match) {
const letter = match[1]
const num = parseInt(match[2], 10)
const newIndex = num < 100 ? `${letter}${num + 1}` : `${String.fromCharCode(letter.charCodeAt(0) + 1)}1`
if (/^[a-z]\d+$/.test(newIndex)) {
editor.updateShape({ id: shapeId, type: shape.type, index: newIndex as any })
}
}
}
} catch (error) {
// Silently fail if shape doesn't exist or operation fails
// This prevents console spam if shape is deleted during selection
}
}
}, [editor, shapeId, isSelected])
// Calculate header background color (lighter shade of primary color) // Calculate header background color (lighter shade of primary color)
const headerBgColor = isSelected const headerBgColor = isSelected
@ -58,13 +154,13 @@ export const StandardizedToolWrapper: React.FC<StandardizedToolWrapperProps> = (
const wrapperStyle: React.CSSProperties = { const wrapperStyle: React.CSSProperties = {
width: typeof width === 'number' ? `${width}px` : width, width: typeof width === 'number' ? `${width}px` : width,
height: isMinimized ? 40 : (typeof height === 'number' ? `${height}px` : height), // Minimized height is just the header height: isMinimized ? 40 : (typeof height === 'number' ? `${height}px` : height), // Minimized height is just the header
backgroundColor: "white", backgroundColor: colors.contentBg,
border: isSelected ? `2px solid ${primaryColor}` : `1px solid ${primaryColor}40`, border: isSelected ? `2px solid ${primaryColor}` : `1px solid ${primaryColor}40`,
borderRadius: "8px", borderRadius: "8px",
overflow: "hidden", overflow: "hidden",
boxShadow: isSelected boxShadow: isSelected
? `0 0 0 2px ${primaryColor}40, 0 4px 8px rgba(0,0,0,0.15)` ? `0 0 0 2px ${primaryColor}40, 0 4px 8px rgba(0,0,0,${isDarkMode ? '0.4' : '0.15'})`
: '0 2px 4px rgba(0,0,0,0.1)', : `0 2px 4px rgba(0,0,0,${isDarkMode ? '0.3' : '0.1'})`,
display: 'flex', display: 'flex',
flexDirection: 'column', flexDirection: 'column',
fontFamily: "Inter, sans-serif", fontFamily: "Inter, sans-serif",
@ -107,8 +203,8 @@ export const StandardizedToolWrapper: React.FC<StandardizedToolWrapperProps> = (
} }
const buttonBaseStyle: React.CSSProperties = { const buttonBaseStyle: React.CSSProperties = {
width: '20px', width: '24px',
height: '20px', height: '24px',
borderRadius: '4px', borderRadius: '4px',
border: 'none', border: 'none',
cursor: 'pointer', cursor: 'pointer',
@ -120,6 +216,9 @@ export const StandardizedToolWrapper: React.FC<StandardizedToolWrapperProps> = (
transition: 'background-color 0.15s ease, color 0.15s ease', transition: 'background-color 0.15s ease, color 0.15s ease',
pointerEvents: 'auto', pointerEvents: 'auto',
flexShrink: 0, flexShrink: 0,
touchAction: 'manipulation', // Prevent double-tap zoom, improve touch responsiveness
padding: 0,
margin: 0,
} }
const minimizeButtonStyle: React.CSSProperties = { const minimizeButtonStyle: React.CSSProperties = {
@ -128,6 +227,16 @@ export const StandardizedToolWrapper: React.FC<StandardizedToolWrapperProps> = (
color: isSelected ? 'white' : primaryColor, color: isSelected ? 'white' : primaryColor,
} }
const pinButtonStyle: React.CSSProperties = {
...buttonBaseStyle,
backgroundColor: isPinnedToView
? (isSelected ? 'rgba(255,255,255,0.4)' : primaryColor)
: (isSelected ? 'rgba(255,255,255,0.2)' : `${primaryColor}20`),
color: isPinnedToView
? (isSelected ? 'white' : 'white')
: (isSelected ? 'white' : primaryColor),
}
const closeButtonStyle: React.CSSProperties = { const closeButtonStyle: React.CSSProperties = {
...buttonBaseStyle, ...buttonBaseStyle,
backgroundColor: isSelected ? 'rgba(255,255,255,0.2)' : `${primaryColor}20`, backgroundColor: isSelected ? 'rgba(255,255,255,0.2)' : `${primaryColor}20`,
@ -143,27 +252,148 @@ export const StandardizedToolWrapper: React.FC<StandardizedToolWrapperProps> = (
transition: 'height 0.2s ease', transition: 'height 0.2s ease',
display: 'flex', display: 'flex',
flexDirection: 'column', flexDirection: 'column',
flex: 1,
} }
const tagsContainerStyle: React.CSSProperties = {
padding: '8px 12px',
borderTop: `1px solid ${colors.tagsBorder}`,
display: 'flex',
flexWrap: 'wrap',
gap: '4px',
alignItems: 'center',
minHeight: '32px',
backgroundColor: colors.tagsBg,
flexShrink: 0,
touchAction: 'manipulation', // Improve touch responsiveness
}
const tagStyle: React.CSSProperties = {
backgroundColor: colors.tagBg,
color: colors.tagText,
padding: '4px 8px', // Increased padding for better touch target
borderRadius: '12px',
fontSize: '10px',
fontWeight: '500',
display: 'inline-flex',
alignItems: 'center',
gap: '4px',
cursor: tagsEditable ? 'pointer' : 'default',
touchAction: 'manipulation', // Improve touch responsiveness
minHeight: '24px', // Ensure adequate touch target height
}
const tagInputStyle: React.CSSProperties = {
border: `1px solid ${colors.inputBorder}`,
borderRadius: '12px',
padding: '2px 6px',
fontSize: '10px',
outline: 'none',
minWidth: '60px',
flex: 1,
backgroundColor: colors.inputBg,
color: isDarkMode ? '#e4e4e4' : '#333',
}
const addTagButtonStyle: React.CSSProperties = {
backgroundColor: colors.addTagBg,
color: colors.tagText,
border: 'none',
borderRadius: '12px',
padding: '4px 10px', // Increased padding for better touch target
fontSize: '10px',
fontWeight: '500',
cursor: 'pointer',
display: 'flex',
alignItems: 'center',
gap: '4px',
touchAction: 'manipulation', // Improve touch responsiveness
minHeight: '24px', // Ensure adequate touch target height
}
const handleTagClick = (tag: string) => {
if (tagsEditable && onTagsChange) {
// Remove tag on click
const newTags = tags.filter(t => t !== tag)
onTagsChange(newTags)
}
}
const handleAddTag = () => {
if (editingTagInput.trim() && onTagsChange) {
const newTag = editingTagInput.trim().replace('#', '')
if (newTag && !tags.includes(newTag) && !tags.includes(`#${newTag}`)) {
const tagToAdd = newTag.startsWith('#') ? newTag : newTag
onTagsChange([...tags, tagToAdd])
}
setEditingTagInput('')
setIsEditingTags(false)
}
}
const handleTagInputKeyDown = (e: React.KeyboardEvent<HTMLInputElement>) => {
if (e.key === 'Enter') {
e.preventDefault()
e.stopPropagation()
handleAddTag()
} else if (e.key === 'Escape') {
e.preventDefault()
e.stopPropagation()
setIsEditingTags(false)
setEditingTagInput('')
} else if (e.key === 'Backspace' && editingTagInput === '' && tags.length > 0) {
// Remove last tag if backspace on empty input
e.stopPropagation()
if (onTagsChange) {
onTagsChange(tags.slice(0, -1))
}
}
}
useEffect(() => {
if (isEditingTags && tagInputRef.current) {
tagInputRef.current.focus()
}
}, [isEditingTags])
const handleHeaderPointerDown = (e: React.PointerEvent) => { const handleHeaderPointerDown = (e: React.PointerEvent) => {
// Check if this is an interactive element (button) // Check if this is an interactive element (button)
const target = e.target as HTMLElement const target = e.target as HTMLElement
const isInteractive = const isInteractive =
target.tagName === 'BUTTON' || target.tagName === 'BUTTON' ||
target.closest('button') || target.closest('button') ||
target.closest('[role="button"]') target.closest('[role="button"]')
if (isInteractive) { if (isInteractive) {
// Buttons handle their own behavior and stop propagation // Buttons handle their own behavior and stop propagation
return return
} }
// Don't stop the event - let tldraw handle it naturally // CRITICAL: Switch to select tool and select this shape when dragging header
// The hand tool override will detect shapes and handle dragging // This ensures dragging works regardless of which tool is currently active
if (editor && shapeId) {
const currentTool = editor.getCurrentToolId()
if (currentTool !== 'select') {
editor.setCurrentTool('select')
}
// Select this shape if not already selected
if (!isSelected) {
editor.setSelectedShapes([shapeId])
}
}
// Don't stop the event - let tldraw handle the drag naturally
} }
const handleButtonClick = (e: React.MouseEvent, action: () => void) => { const handleButtonClick = (e: React.MouseEvent, action: () => void) => {
e.stopPropagation() e.stopPropagation()
e.preventDefault()
action()
}
const handleButtonTouch = (e: React.TouchEvent, action: () => void) => {
e.stopPropagation()
e.preventDefault()
action() action()
} }
@ -197,7 +427,18 @@ export const StandardizedToolWrapper: React.FC<StandardizedToolWrapperProps> = (
onPointerDown={handleHeaderPointerDown} onPointerDown={handleHeaderPointerDown}
onMouseEnter={() => setIsHoveringHeader(true)} onMouseEnter={() => setIsHoveringHeader(true)}
onMouseLeave={() => setIsHoveringHeader(false)} onMouseLeave={() => setIsHoveringHeader(false)}
onMouseDown={(_e) => { onMouseDown={(e) => {
// Don't select if clicking on a button - let the button handle the click
const target = e.target as HTMLElement
const isButton =
target.tagName === 'BUTTON' ||
target.closest('button') ||
target.closest('[role="button"]')
if (isButton) {
return
}
// Ensure selection happens on mouse down for immediate visual feedback // Ensure selection happens on mouse down for immediate visual feedback
if (editor && shapeId && !isSelected) { if (editor && shapeId && !isSelected) {
editor.setSelectedShapes([shapeId]) editor.setSelectedShapes([shapeId])
@ -209,6 +450,20 @@ export const StandardizedToolWrapper: React.FC<StandardizedToolWrapperProps> = (
{headerContent || title} {headerContent || title}
</div> </div>
<div style={buttonContainerStyle}> <div style={buttonContainerStyle}>
{onPinToggle && (
<button
style={pinButtonStyle}
onClick={(e) => handleButtonClick(e, onPinToggle)}
onPointerDown={(e) => e.stopPropagation()}
onMouseDown={(e) => e.stopPropagation()}
onTouchStart={(e) => handleButtonTouch(e, onPinToggle)}
onTouchEnd={(e) => e.stopPropagation()}
title={isPinnedToView ? "Unpin from view" : "Pin to view"}
aria-label={isPinnedToView ? "Unpin from view" : "Pin to view"}
>
📌
</button>
)}
<button <button
style={minimizeButtonStyle} style={minimizeButtonStyle}
onClick={(e) => { onClick={(e) => {
@ -220,6 +475,13 @@ export const StandardizedToolWrapper: React.FC<StandardizedToolWrapperProps> = (
} }
}} }}
onPointerDown={(e) => e.stopPropagation()} onPointerDown={(e) => e.stopPropagation()}
onMouseDown={(e) => e.stopPropagation()}
onTouchStart={(e) => {
if (onMinimize) {
handleButtonTouch(e, onMinimize)
}
}}
onTouchEnd={(e) => e.stopPropagation()}
title="Minimize" title="Minimize"
aria-label="Minimize" aria-label="Minimize"
disabled={!onMinimize} disabled={!onMinimize}
@ -230,6 +492,9 @@ export const StandardizedToolWrapper: React.FC<StandardizedToolWrapperProps> = (
style={closeButtonStyle} style={closeButtonStyle}
onClick={(e) => handleButtonClick(e, onClose)} onClick={(e) => handleButtonClick(e, onClose)}
onPointerDown={(e) => e.stopPropagation()} onPointerDown={(e) => e.stopPropagation()}
onMouseDown={(e) => e.stopPropagation()}
onTouchStart={(e) => handleButtonTouch(e, onClose)}
onTouchEnd={(e) => e.stopPropagation()}
title="Close" title="Close"
aria-label="Close" aria-label="Close"
> >
@ -240,12 +505,87 @@ export const StandardizedToolWrapper: React.FC<StandardizedToolWrapperProps> = (
{/* Content Area */} {/* Content Area */}
{!isMinimized && ( {!isMinimized && (
<div <>
style={contentStyle} <div
onPointerDown={handleContentPointerDown} style={contentStyle}
> onPointerDown={handleContentPointerDown}
{children} >
</div> {children}
</div>
{/* Tags at the bottom */}
{(tags.length > 0 || (tagsEditable && isSelected)) && (
<div
style={tagsContainerStyle}
onPointerDown={(e) => e.stopPropagation()}
onTouchStart={(e) => e.stopPropagation()}
onClick={(e) => {
if (tagsEditable && !isEditingTags && e.target === e.currentTarget) {
setIsEditingTags(true)
}
}}
>
{tags.slice(0, 5).map((tag, index) => (
<span
key={index}
style={tagStyle}
onClick={(e) => {
e.stopPropagation()
handleTagClick(tag)
}}
onTouchEnd={(e) => {
e.stopPropagation()
e.preventDefault()
handleTagClick(tag)
}}
title={tagsEditable ? "Click to remove tag" : undefined}
>
{tag.replace('#', '')}
{tagsEditable && <span style={{ fontSize: '8px' }}>×</span>}
</span>
))}
{tags.length > 5 && (
<span style={tagStyle}>
+{tags.length - 5}
</span>
)}
{isEditingTags && (
<input
ref={tagInputRef}
type="text"
value={editingTagInput}
onChange={(e) => setEditingTagInput(e.target.value)}
onKeyDown={handleTagInputKeyDown}
onBlur={() => {
handleAddTag()
}}
style={tagInputStyle}
placeholder="Add tag..."
onPointerDown={(e) => e.stopPropagation()}
/>
)}
{!isEditingTags && tagsEditable && isSelected && tags.length < 10 && (
<button
style={addTagButtonStyle}
onClick={(e) => {
e.stopPropagation()
setIsEditingTags(true)
}}
onPointerDown={(e) => e.stopPropagation()}
onTouchStart={(e) => {
e.stopPropagation()
e.preventDefault()
setIsEditingTags(true)
}}
onTouchEnd={(e) => e.stopPropagation()}
title="Add tag"
>
+ Add
</button>
)}
</div>
)}
</>
)} )}
</div> </div>
) )

16
src/components/StarBoardButton.tsx
View File

@ -85,15 +85,21 @@ const StarBoardButton: React.FC<StarBoardButtonProps> = ({ className = '' }) =>
<button <button
onClick={handleStarToggle} onClick={handleStarToggle}
disabled={isLoading} disabled={isLoading}
className={`star-board-button ${className} ${isStarred ? 'starred' : ''}`} className={`toolbar-btn star-board-button ${className} ${isStarred ? 'starred' : ''}`}
title={isStarred ? 'Remove from starred boards' : 'Add to starred boards'} title={isStarred ? 'Remove from starred boards' : 'Add to starred boards'}
> >
{isLoading ? ( {isLoading ? (
<span className="loading-spinner">⏳</span> <svg width="14" height="14" viewBox="0 0 16 16" fill="currentColor" className="loading-spinner">
) : isStarred ? ( <path d="M8 3a5 5 0 1 0 4.546 2.914.5.5 0 0 1 .908-.417A6 6 0 1 1 8 2v1z"/>
<span className="star-icon starred">⭐</span> </svg>
) : ( ) : (
<span className="star-icon">☆</span> <svg width="14" height="14" viewBox="0 0 16 16" fill="currentColor">
{isStarred ? (
<path d="M3.612 15.443c-.386.198-.824-.149-.746-.592l.83-4.73L.173 6.765c-.329-.314-.158-.888.283-.95l4.898-.696L7.538.792c.197-.39.73-.39.927 0l2.184 4.327 4.898.696c.441.062.612.636.282.95l-3.522 3.356.83 4.73c.078.443-.36.79-.746.592L8 13.187l-4.389 2.256z"/>
) : (
<path d="M2.866 14.85c-.078.444.36.791.746.593l4.39-2.256 4.389 2.256c.386.198.824-.149.746-.592l-.83-4.73 3.522-3.356c.33-.314.16-.888-.282-.95l-4.898-.696L8.465.792a.513.513 0 0 0-.927 0L5.354 5.12l-4.898.696c-.441.062-.612.636-.283.95l3.523 3.356-.83 4.73zm4.905-2.767-3.686 1.894.694-3.957a.565.565 0 0 0-.163-.505L1.71 6.745l4.052-.576a.525.525 0 0 0 .393-.288L8 2.223l1.847 3.658a.525.525 0 0 0 .393.288l4.052.575-2.906 2.77a.565.565 0 0 0-.163.506l.694 3.957-3.686-1.894a.503.503 0 0 0-.461 0z"/>
)}
</svg>
)} )}
</button> </button>

18
src/components/auth/CryptoLogin.tsx → src/components/auth/CryptID.tsx
View File

@ -4,15 +4,15 @@ import { useAuth } from '../../context/AuthContext';
import { useNotifications } from '../../context/NotificationContext'; import { useNotifications } from '../../context/NotificationContext';
import { checkBrowserSupport, isSecureContext } from '../../lib/utils/browser'; import { checkBrowserSupport, isSecureContext } from '../../lib/utils/browser';
interface CryptoLoginProps { interface CryptIDProps {
onSuccess?: () => void; onSuccess?: () => void;
onCancel?: () => void; onCancel?: () => void;
} }
/** /**
* WebCryptoAPI-based authentication component * CryptID - WebCryptoAPI-based authentication component
*/ */
const CryptoLogin: React.FC<CryptoLoginProps> = ({ onSuccess, onCancel }) => { const CryptID: React.FC<CryptIDProps> = ({ onSuccess, onCancel }) => {
const [username, setUsername] = useState(''); const [username, setUsername] = useState('');
const [isRegistering, setIsRegistering] = useState(false); const [isRegistering, setIsRegistering] = useState(false);
const [error, setError] = useState<string | null>(null); const [error, setError] = useState<string | null>(null);
@ -178,7 +178,7 @@ const CryptoLogin: React.FC<CryptoLoginProps> = ({ onSuccess, onCancel }) => {
return ( return (
<div className="crypto-login-container"> <div className="crypto-login-container">
<h2>{isRegistering ? 'Create Cryptographic Account' : 'Cryptographic Sign In'}</h2> <h2>{isRegistering ? 'Create CryptID Account' : 'CryptID Sign In'}</h2>
{/* Show existing users if available */} {/* Show existing users if available */}
{existingUsers.length > 0 && !isRegistering && ( {existingUsers.length > 0 && !isRegistering && (
@ -206,11 +206,11 @@ const CryptoLogin: React.FC<CryptoLoginProps> = ({ onSuccess, onCancel }) => {
<div className="crypto-info"> <div className="crypto-info">
<p> <p>
{isRegistering {isRegistering
? 'Create a new account using WebCryptoAPI for secure authentication.' ? 'Create a new CryptID account using WebCryptoAPI for secure authentication.'
: existingUsers.length > 0 : existingUsers.length > 0
? 'Select an account above or enter a different username to sign in.' ? 'Select an account above or enter a different username to sign in.'
: 'Sign in using your cryptographic credentials.' : 'Sign in using your CryptID credentials.'
} }
</p> </p>
<div className="crypto-features"> <div className="crypto-features">
@ -276,4 +276,4 @@ const CryptoLogin: React.FC<CryptoLoginProps> = ({ onSuccess, onCancel }) => {
); );
}; };
export default CryptoLogin; export default CryptID;

2
src/components/auth/CryptoDebug.tsx
View File

@ -145,7 +145,7 @@ const CryptoDebug: React.FC = () => {
const storedUsers = JSON.parse(localStorage.getItem('registeredUsers') || '[]'); const storedUsers = JSON.parse(localStorage.getItem('registeredUsers') || '[]');
addResult(`All registered users: ${JSON.stringify(storedUsers)}`); addResult(`All registered users: ${JSON.stringify(storedUsers)}`);
// Filter for users with valid keys (same logic as CryptoLogin) // Filter for users with valid keys (same logic as CryptID)
const validUsers = storedUsers.filter((user: string) => { const validUsers = storedUsers.filter((user: string) => {
const publicKey = localStorage.getItem(`${user}_publicKey`); const publicKey = localStorage.getItem(`${user}_publicKey`);
if (!publicKey) return false; if (!publicKey) return false;

102
src/components/auth/LinkDevice.tsx
View File

@ -1,102 +0,0 @@
import React, { useState, useEffect } from 'react'
import { useNavigate } from 'react-router-dom'
import { createAccountLinkingConsumer } from '../../lib/auth/linking'
import { useAuth } from '../../context/AuthContext'
import { useNotifications } from '../../context/NotificationContext'
const LinkDevice: React.FC = () => {
const [username, setUsername] = useState('')
const [displayPin, setDisplayPin] = useState('')
const [view, setView] = useState<'enter-username' | 'show-pin' | 'load-filesystem'>('enter-username')
const [accountLinkingConsumer, setAccountLinkingConsumer] = useState<any>(null)
const navigate = useNavigate()
const { login } = useAuth()
const { addNotification } = useNotifications()
const initAccountLinkingConsumer = async () => {
try {
const consumer = await createAccountLinkingConsumer(username)
setAccountLinkingConsumer(consumer)
consumer.on('challenge', ({ pin }: { pin: number[] }) => {
setDisplayPin(pin.join(''))
setView('show-pin')
})
consumer.on('link', async ({ approved, username }: { approved: boolean, username: string }) => {
if (approved) {
setView('load-filesystem')
const success = await login(username)
if (success) {
addNotification("You're now connected!", "success")
navigate('/')
} else {
addNotification("Connection successful but login failed", "error")
navigate('/login')
}
} else {
addNotification('The connection attempt was cancelled', "warning")
navigate('/')
}
})
} catch (error) {
console.error('Error initializing account linking consumer:', error)
addNotification('Failed to initialize device linking', "error")
}
}
const handleSubmitUsername = (e: React.FormEvent) => {
e.preventDefault()
initAccountLinkingConsumer()
}
// Clean up consumer on unmount
useEffect(() => {
return () => {
if (accountLinkingConsumer) {
accountLinkingConsumer.destroy()
}
}
}, [accountLinkingConsumer])
return (
<div className="link-device-container">
{view === 'enter-username' && (
<>
<h2>Link a New Device</h2>
<form onSubmit={handleSubmitUsername}>
<div className="form-group">
<label htmlFor="username">Username</label>
<input
type="text"
id="username"
value={username}
onChange={(e) => setUsername(e.target.value)}
required
/>
</div>
<button type="submit" disabled={!username}>Continue</button>
</form>
</>
)}
{view === 'show-pin' && (
<div className="pin-display">
<h2>Enter this PIN on your other device</h2>
<div className="pin-code">{displayPin}</div>
</div>
)}
{view === 'load-filesystem' && (
<div className="loading">
<h2>Loading your filesystem...</h2>
<p>Please wait while we connect to your account.</p>
</div>
)}
</div>
)
}
export default LinkDevice

12
src/components/auth/LoginButton.tsx
View File

@ -1,7 +1,7 @@
import React, { useState } from 'react'; import React, { useState } from 'react';
import { useAuth } from '../../context/AuthContext'; import { useAuth } from '../../context/AuthContext';
import { useNotifications } from '../../context/NotificationContext'; import { useNotifications } from '../../context/NotificationContext';
import CryptoLogin from './CryptoLogin'; import CryptID from './CryptID';
interface LoginButtonProps { interface LoginButtonProps {
className?: string; className?: string;
@ -33,16 +33,20 @@ const LoginButton: React.FC<LoginButtonProps> = ({ className = '' }) => {
<> <>
<button <button
onClick={handleLoginClick} onClick={handleLoginClick}
className={`login-button ${className}`} className={`toolbar-btn login-button ${className}`}
title="Sign in to save your work and access additional features" title="Sign in to save your work and access additional features"
> >
Sign In <svg width="14" height="14" viewBox="0 0 16 16" fill="currentColor">
<path fillRule="evenodd" d="M6 3.5a.5.5 0 0 1 .5-.5h8a.5.5 0 0 1 .5.5v9a.5.5 0 0 1-.5.5h-8a.5.5 0 0 1-.5-.5v-2a.5.5 0 0 0-1 0v2A1.5 1.5 0 0 0 6.5 14h8a1.5 1.5 0 0 0 1.5-1.5v-9A1.5 1.5 0 0 0 14.5 2h-8A1.5 1.5 0 0 0 5 3.5v2a.5.5 0 0 0 1 0v-2z"/>
<path fillRule="evenodd" d="M11.854 8.354a.5.5 0 0 0 0-.708l-3-3a.5.5 0 1 0-.708.708L10.293 7.5H1.5a.5.5 0 0 0 0 1h8.793l-2.147 2.146a.5.5 0 0 0 .708.708l3-3z"/>
</svg>
<span>Sign In</span>
</button> </button>
{showLogin && ( {showLogin && (
<div className="login-overlay"> <div className="login-overlay">
<div className="login-modal"> <div className="login-modal">
<CryptoLogin <CryptID
onSuccess={handleLoginSuccess} onSuccess={handleLoginSuccess}
onCancel={handleLoginCancel} onCancel={handleLoginCancel}
/> />

2
src/components/auth/Profile.tsx
View File

@ -63,7 +63,7 @@ export const Profile: React.FC<ProfileProps> = ({ onLogout, onOpenVaultBrowser }
return ( return (
<div className="profile-container"> <div className="profile-container">
<div className="profile-header"> <div className="profile-header">
<h3>Welcome, {session.username}!</h3> <h3>CryptID: {session.username}</h3>
</div> </div>
<div className="profile-settings"> <div className="profile-settings">

64
src/components/auth/Register.tsx
View File

@ -1,64 +0,0 @@
import React, { useState } from 'react'
import { register } from '../../lib/auth/account'
const Register: React.FC = () => {
const [username, setUsername] = useState('')
const [checkingUsername, setCheckingUsername] = useState(false)
const [initializingFilesystem, setInitializingFilesystem] = useState(false)
const [error, setError] = useState<string | null>(null)
const handleRegister = async (e: React.FormEvent) => {
e.preventDefault()
if (checkingUsername) {
return
}
setInitializingFilesystem(true)
setError(null)
try {
const success = await register(username)
if (!success) {
setError('Registration failed. Username may be taken.')
setInitializingFilesystem(false)
}
} catch (err) {
setError('An error occurred during registration')
setInitializingFilesystem(false)
console.error(err)
}
}
return (
<div className="register-container">
<h2>Create an Account</h2>
<form onSubmit={handleRegister}>
<div className="form-group">
<label htmlFor="username">Username</label>
<input
type="text"
id="username"
value={username}
onChange={(e) => setUsername(e.target.value)}
disabled={initializingFilesystem}
required
/>
</div>
{error && <div className="error-message">{error}</div>}
<button
type="submit"
disabled={initializingFilesystem || !username}
>
{initializingFilesystem ? 'Creating Account...' : 'Create Account'}
</button>
</form>
</div>
)
}
export default Register

187
src/components/location/LocationCapture.tsx
View File

@ -1,187 +0,0 @@
"use client"
import type React from "react"
import { useState, useEffect } from "react"
import { useAuth } from "@/context/AuthContext"
import { LocationStorageService, type LocationData } from "@/lib/location/locationStorage"
import type { GeolocationPosition } from "@/lib/location/types"
interface LocationCaptureProps {
onLocationCaptured?: (location: LocationData) => void
onError?: (error: string) => void
}
export const LocationCapture: React.FC<LocationCaptureProps> = ({ onLocationCaptured, onError }) => {
const { session, fileSystem } = useAuth()
const [isCapturing, setIsCapturing] = useState(false)
const [permissionState, setPermissionState] = useState<"prompt" | "granted" | "denied">("prompt")
const [currentLocation, setCurrentLocation] = useState<GeolocationPosition | null>(null)
const [error, setError] = useState<string | null>(null)
// Show loading state while auth is initializing
if (session.loading) {
return (
<div className="location-capture-loading flex items-center justify-center min-h-[200px]">
<div className="text-center">
<div className="text-2xl mb-2 animate-spin">⏳</div>
<p className="text-sm text-muted-foreground">Loading authentication...</p>
</div>
</div>
)
}
// Check permission status on mount
useEffect(() => {
if ("permissions" in navigator) {
navigator.permissions.query({ name: "geolocation" }).then((result) => {
setPermissionState(result.state as "prompt" | "granted" | "denied")
result.addEventListener("change", () => {
setPermissionState(result.state as "prompt" | "granted" | "denied")
})
})
}
}, [])
const captureLocation = async () => {
// Don't proceed if still loading
if (session.loading) {
return
}
if (!session.authed) {
const errorMsg = "You must be logged in to share your location. Please log in and try again."
setError(errorMsg)
onError?.(errorMsg)
return
}
if (!fileSystem) {
const errorMsg = "File system not available. Please refresh the page and try again."
setError(errorMsg)
onError?.(errorMsg)
return
}
setIsCapturing(true)
setError(null)
try {
// Request geolocation
const position = await new Promise<GeolocationPosition>((resolve, reject) => {
navigator.geolocation.getCurrentPosition(
(pos) => resolve(pos as GeolocationPosition),
(err) => reject(err),
{
enableHighAccuracy: true,
timeout: 10000,
maximumAge: 0,
},
)
})
setCurrentLocation(position)
// Create location data
const locationData: LocationData = {
id: crypto.randomUUID(),
userId: session.username,
latitude: position.coords.latitude,
longitude: position.coords.longitude,
accuracy: position.coords.accuracy,
timestamp: position.timestamp,
expiresAt: null, // Will be set when creating a share
precision: "exact",
}
// Save to filesystem
const storageService = new LocationStorageService(fileSystem)
await storageService.initialize()
await storageService.saveLocation(locationData)
onLocationCaptured?.(locationData)
} catch (err: any) {
let errorMsg = "Failed to capture location"
if (err.code === 1) {
errorMsg = "Location permission denied. Please enable location access in your browser settings."
setPermissionState("denied")
} else if (err.code === 2) {
errorMsg = "Location unavailable. Please check your device settings."
} else if (err.code === 3) {
errorMsg = "Location request timed out. Please try again."
}
setError(errorMsg)
onError?.(errorMsg)
} finally {
setIsCapturing(false)
}
}
return (
<div className="location-capture">
<div className="capture-header">
<h2 className="text-2xl font-semibold text-balance">Share Your Location</h2>
<p className="text-sm text-muted-foreground mt-2">Securely share your current location with others</p>
</div>
{/* Permission status */}
{permissionState === "denied" && (
<div className="permission-denied bg-destructive/10 border border-destructive/20 rounded-lg p-4 mt-4">
<p className="text-sm text-destructive">
Location access is blocked. Please enable it in your browser settings to continue.
</p>
</div>
)}
{/* Current location display */}
{currentLocation && (
<div className="current-location bg-muted/50 rounded-lg p-4 mt-4">
<h3 className="text-sm font-medium mb-2">Current Location</h3>
<div className="location-details text-xs space-y-1">
<p>
<span className="text-muted-foreground">Latitude:</span> {currentLocation.coords.latitude.toFixed(6)}
</p>
<p>
<span className="text-muted-foreground">Longitude:</span> {currentLocation.coords.longitude.toFixed(6)}
</p>
<p>
<span className="text-muted-foreground">Accuracy:</span> ±{Math.round(currentLocation.coords.accuracy)}m
</p>
<p className="text-muted-foreground">Captured {new Date(currentLocation.timestamp).toLocaleString()}</p>
</div>
</div>
)}
{/* Error display */}
{error && (
<div className="error-message bg-destructive/10 border border-destructive/20 rounded-lg p-4 mt-4">
<p className="text-sm text-destructive">{error}</p>
</div>
)}
{/* Capture button */}
<button
onClick={captureLocation}
disabled={isCapturing || permissionState === "denied" || !session.authed}
className="capture-button w-full mt-6 bg-primary text-primary-foreground hover:bg-primary/90 disabled:opacity-50 disabled:cursor-not-allowed rounded-lg px-6 py-3 font-medium transition-colors"
>
{isCapturing ? (
<span className="flex items-center justify-center gap-2">
<span className="spinner" />
Capturing Location...
</span>
) : (
"Capture My Location"
)}
</button>
{!session.authed && (
<p className="text-xs text-muted-foreground text-center mt-3">Please log in to share your location</p>
)}
</div>
)
}

270
src/components/location/LocationDashboard.tsx
View File

@ -1,270 +0,0 @@
"use client"
import type React from "react"
import { useState, useEffect } from "react"
import { useAuth } from "@/context/AuthContext"
import { LocationStorageService, type LocationData, type LocationShare } from "@/lib/location/locationStorage"
import { LocationMap } from "./LocationMap"
interface ShareWithLocation {
share: LocationShare
location: LocationData
}
export const LocationDashboard: React.FC = () => {
const { session, fileSystem } = useAuth()
const [shares, setShares] = useState<ShareWithLocation[]>([])
const [loading, setLoading] = useState(true)
const [selectedShare, setSelectedShare] = useState<ShareWithLocation | null>(null)
const [error, setError] = useState<string | null>(null)
const loadShares = async () => {
if (!fileSystem) {
setError("File system not available")
setLoading(false)
return
}
try {
const storageService = new LocationStorageService(fileSystem)
await storageService.initialize()
// Get all shares
const allShares = await storageService.getAllShares()
// Get locations for each share
const sharesWithLocations: ShareWithLocation[] = []
for (const share of allShares) {
const location = await storageService.getLocation(share.locationId)
if (location) {
sharesWithLocations.push({ share, location })
}
}
// Sort by creation date (newest first)
sharesWithLocations.sort((a, b) => b.share.createdAt - a.share.createdAt)
setShares(sharesWithLocations)
setLoading(false)
} catch (err) {
console.error("Error loading shares:", err)
setError("Failed to load location shares")
setLoading(false)
}
}
useEffect(() => {
if (session.authed && fileSystem) {
loadShares()
}
}, [session.authed, fileSystem])
const handleCopyLink = async (shareToken: string) => {
const baseUrl = window.location.origin
const link = `${baseUrl}/location/${shareToken}`
try {
await navigator.clipboard.writeText(link)
alert("Link copied to clipboard!")
} catch (err) {
console.error("Failed to copy link:", err)
alert("Failed to copy link")
}
}
const isExpired = (share: LocationShare): boolean => {
return share.expiresAt ? share.expiresAt < Date.now() : false
}
const isMaxViewsReached = (share: LocationShare): boolean => {
return share.maxViews ? share.viewCount >= share.maxViews : false
}
const getShareStatus = (share: LocationShare): { label: string; color: string } => {
if (isExpired(share)) {
return { label: "Expired", color: "text-destructive" }
}
if (isMaxViewsReached(share)) {
return { label: "Max Views Reached", color: "text-destructive" }
}
return { label: "Active", color: "text-green-600" }
}
if (!session.authed) {
return (
<div className="location-dashboard-auth flex items-center justify-center min-h-[400px]">
<div className="text-center max-w-md">
<div className="text-4xl mb-4">🔒</div>
<h2 className="text-xl font-semibold mb-2">Authentication Required</h2>
<p className="text-sm text-muted-foreground">Please log in to view your location shares</p>
</div>
</div>
)
}
if (loading) {
return (
<div className="location-dashboard flex items-center justify-center min-h-[400px]">
<div className="flex flex-col items-center gap-3">
<div className="spinner" />
<p className="text-sm text-muted-foreground">Loading your shares...</p>
</div>
</div>
)
}
if (error) {
return (
<div className="location-dashboard flex items-center justify-center min-h-[400px]">
<div className="text-center max-w-md">
<div className="text-4xl mb-4">⚠️</div>
<h2 className="text-xl font-semibold mb-2">Error Loading Dashboard</h2>
<p className="text-sm text-muted-foreground">{error}</p>
<button
onClick={loadShares}
className="mt-4 px-6 py-2 rounded-lg bg-primary text-primary-foreground hover:bg-primary/90 transition-colors"
>
Retry
</button>
</div>
</div>
)
}
return (
<div className="location-dashboard max-w-6xl mx-auto p-6">
<div className="dashboard-header mb-8">
<h1 className="text-3xl font-bold text-balance">Location Shares</h1>
<p className="text-sm text-muted-foreground mt-2">Manage your shared locations and privacy settings</p>
</div>
{shares.length === 0 ? (
<div className="empty-state flex flex-col items-center justify-center min-h-[400px] text-center">
<div className="text-6xl mb-4">📍</div>
<h2 className="text-xl font-semibold mb-2">No Location Shares Yet</h2>
<p className="text-sm text-muted-foreground mb-6 max-w-md">
You haven't shared any locations yet. Create your first share to get started.
</p>
<a
href="/share-location"
className="px-6 py-3 rounded-lg bg-primary text-primary-foreground hover:bg-primary/90 transition-colors font-medium"
>
Share Your Location
</a>
</div>
) : (
<div className="dashboard-content">
{/* Stats Overview */}
<div className="stats-grid grid grid-cols-1 md:grid-cols-3 gap-4 mb-8">
<div className="stat-card bg-muted/50 rounded-lg p-4 border border-border">
<div className="stat-label text-sm text-muted-foreground mb-1">Total Shares</div>
<div className="stat-value text-3xl font-bold">{shares.length}</div>
</div>
<div className="stat-card bg-muted/50 rounded-lg p-4 border border-border">
<div className="stat-label text-sm text-muted-foreground mb-1">Active Shares</div>
<div className="stat-value text-3xl font-bold text-green-600">
{shares.filter((s) => !isExpired(s.share) && !isMaxViewsReached(s.share)).length}
</div>
</div>
<div className="stat-card bg-muted/50 rounded-lg p-4 border border-border">
<div className="stat-label text-sm text-muted-foreground mb-1">Total Views</div>
<div className="stat-value text-3xl font-bold">
{shares.reduce((sum, s) => sum + s.share.viewCount, 0)}
</div>
</div>
</div>
{/* Shares List */}
<div className="shares-list space-y-4">
{shares.map(({ share, location }) => {
const status = getShareStatus(share)
const isSelected = selectedShare?.share.id === share.id
return (
<div
key={share.id}
className={`share-card bg-background rounded-lg border-2 transition-colors ${
isSelected ? "border-primary" : "border-border hover:border-primary/50"
}`}
>
<div className="share-card-header p-4 flex items-start justify-between gap-4">
<div className="share-info flex-1">
<div className="flex items-center gap-3 mb-2">
<h3 className="font-semibold">Location Share</h3>
<span className={`text-xs font-medium ${status.color}`}>{status.label}</span>
</div>
<div className="share-meta text-xs text-muted-foreground space-y-1">
<p>Created: {new Date(share.createdAt).toLocaleString()}</p>
{share.expiresAt && <p>Expires: {new Date(share.expiresAt).toLocaleString()}</p>}
<p>
Views: {share.viewCount}
{share.maxViews && ` / ${share.maxViews}`}
</p>
<p>
Precision: <span className="capitalize">{share.precision}</span>
</p>
</div>
</div>
<div className="share-actions flex gap-2">
<button
onClick={() => handleCopyLink(share.shareToken)}
disabled={isExpired(share) || isMaxViewsReached(share)}
className="px-4 py-2 rounded-lg border border-border hover:bg-muted disabled:opacity-50 disabled:cursor-not-allowed transition-colors text-sm"
>
Copy Link
</button>
<button
onClick={() => setSelectedShare(isSelected ? null : { share, location })}
className="px-4 py-2 rounded-lg bg-primary text-primary-foreground hover:bg-primary/90 transition-colors text-sm"
>
{isSelected ? "Hide" : "View"} Map
</button>
</div>
</div>
{isSelected && (
<div className="share-card-body p-4 pt-0 border-t border-border mt-4">
<LocationMap location={location} precision={share.precision} showAccuracy={true} height="300px" />
</div>
)}
</div>
)
})}
</div>
</div>
)}
</div>
)
}

241
src/components/location/LocationMap.tsx
View File

@ -1,241 +0,0 @@
"use client"
import type React from "react"
import { useEffect, useRef, useState } from "react"
import type { LocationData } from "@/lib/location/locationStorage"
import { obfuscateLocation } from "@/lib/location/locationStorage"
import type { PrecisionLevel } from "@/lib/location/types"
// Leaflet types
interface LeafletMap {
setView: (coords: [number, number], zoom: number) => void
remove: () => void
}
interface LeafletMarker {
addTo: (map: LeafletMap) => LeafletMarker
bindPopup: (content: string) => LeafletMarker
}
interface LeafletCircle {
addTo: (map: LeafletMap) => LeafletCircle
}
interface LeafletTileLayer {
addTo: (map: LeafletMap) => LeafletTileLayer
}
interface Leaflet {
map: (element: HTMLElement, options?: any) => LeafletMap
marker: (coords: [number, number], options?: any) => LeafletMarker
circle: (coords: [number, number], options?: any) => LeafletCircle
tileLayer: (url: string, options?: any) => LeafletTileLayer
icon: (options: any) => any
}
declare global {
interface Window {
L?: Leaflet
}
}
interface LocationMapProps {
location: LocationData
precision?: PrecisionLevel
showAccuracy?: boolean
height?: string
}
export const LocationMap: React.FC<LocationMapProps> = ({
location,
precision = "exact",
showAccuracy = true,
height = "400px",
}) => {
const mapContainer = useRef<HTMLDivElement>(null)
const mapInstance = useRef<LeafletMap | null>(null)
const [isLoading, setIsLoading] = useState(true)
const [error, setError] = useState<string | null>(null)
useEffect(() => {
// Load Leaflet CSS and JS
const loadLeaflet = async () => {
try {
// Load CSS
if (!document.querySelector('link[href*="leaflet.css"]')) {
const link = document.createElement("link")
link.rel = "stylesheet"
link.href = "https://unpkg.com/leaflet@1.9.4/dist/leaflet.css"
link.integrity = "sha256-p4NxAoJBhIIN+hmNHrzRCf9tD/miZyoHS5obTRR9BMY="
link.crossOrigin = ""
document.head.appendChild(link)
}
// Load JS
if (!window.L) {
await new Promise<void>((resolve, reject) => {
const script = document.createElement("script")
script.src = "https://unpkg.com/leaflet@1.9.4/dist/leaflet.js"
script.integrity = "sha256-20nQCchB9co0qIjJZRGuk2/Z9VM+kNiyxNV1lvTlZBo="
script.crossOrigin = ""
script.onload = () => resolve()
script.onerror = () => reject(new Error("Failed to load Leaflet"))
document.head.appendChild(script)
})
}
setIsLoading(false)
} catch (err) {
setError("Failed to load map library")
setIsLoading(false)
}
}
loadLeaflet()
}, [])
useEffect(() => {
if (!mapContainer.current || !window.L || isLoading) return
// Clean up existing map
if (mapInstance.current) {
mapInstance.current.remove()
}
const L = window.L!
// Get obfuscated location based on precision
const { lat, lng, radius } = obfuscateLocation(location.latitude, location.longitude, precision)
// Create map
const map = L.map(mapContainer.current, {
center: [lat, lng],
zoom: precision === "exact" ? 15 : precision === "street" ? 14 : precision === "neighborhood" ? 12 : 10,
zoomControl: true,
attributionControl: true,
})
// Add OpenStreetMap tiles
L.tileLayer("https://{s}.tile.openstreetmap.org/{z}/{x}/{y}.png", {
attribution: '&copy; <a href="https://www.openstreetmap.org/copyright">OpenStreetMap</a> contributors',
maxZoom: 19,
}).addTo(map)
// Add marker
const marker = L.marker([lat, lng], {
icon: L.icon({
iconUrl: "https://unpkg.com/leaflet@1.9.4/dist/images/marker-icon.png",
iconRetinaUrl: "https://unpkg.com/leaflet@1.9.4/dist/images/marker-icon-2x.png",
shadowUrl: "https://unpkg.com/leaflet@1.9.4/dist/images/marker-shadow.png",
iconSize: [25, 41],
iconAnchor: [12, 41],
popupAnchor: [1, -34],
shadowSize: [41, 41],
}),
}).addTo(map)
// Add popup with location info
const popupContent = `
<div style="font-family: system-ui, sans-serif;">
<strong>Shared Location</strong><br/>
<small style="color: #666;">
Precision: ${precision}<br/>
${new Date(location.timestamp).toLocaleString()}
</small>
</div>
`
marker.bindPopup(popupContent)
// Add accuracy circle if showing accuracy
if (showAccuracy && radius > 0) {
L.circle([lat, lng], {
radius: radius,
color: "#3b82f6",
fillColor: "#3b82f6",
fillOpacity: 0.1,
weight: 2,
}).addTo(map)
}
mapInstance.current = map
// Cleanup
return () => {
if (mapInstance.current) {
mapInstance.current.remove()
mapInstance.current = null
}
}
}, [location, precision, showAccuracy, isLoading])
if (error) {
return (
<div
className="map-error flex items-center justify-center bg-muted/50 rounded-lg border border-border"
style={{ height }}
>
<p className="text-sm text-destructive">{error}</p>
</div>
)
}
if (isLoading) {
return (
<div
className="map-loading flex items-center justify-center bg-muted/50 rounded-lg border border-border"
style={{ height }}
>
<div className="flex flex-col items-center gap-3">
<div className="spinner" />
<p className="text-sm text-muted-foreground">Loading map...</p>
</div>
</div>
)
}
return (
<div className="location-map-wrapper">
<div
ref={mapContainer}
className="location-map rounded-lg border border-border overflow-hidden"
style={{ height, width: "100%" }}
/>
<div className="map-info mt-3 text-xs text-muted-foreground">
<p>
Showing {precision} location • Last updated {new Date(location.timestamp).toLocaleTimeString()}
</p>
</div>
</div>
)
}

45
src/components/location/LocationShareDialog.tsx
View File

@ -1,45 +0,0 @@
import {
TLUiDialogProps,
TldrawUiDialogBody,
TldrawUiDialogCloseButton,
TldrawUiDialogHeader,
TldrawUiDialogTitle,
} from "tldraw"
import React from "react"
import { ShareLocation } from "./ShareLocation"
export function LocationShareDialog({ onClose: _onClose }: TLUiDialogProps) {
return (
<>
<TldrawUiDialogHeader>
<TldrawUiDialogTitle>Share Location</TldrawUiDialogTitle>
<TldrawUiDialogCloseButton />
</TldrawUiDialogHeader>
<TldrawUiDialogBody style={{ maxWidth: 800, maxHeight: "90vh", overflow: "auto" }}>
<ShareLocation />
</TldrawUiDialogBody>
</>
)
}

183
src/components/location/LocationViewer.tsx
View File

@ -1,183 +0,0 @@
"use client"
import type React from "react"
import { useState, useEffect } from "react"
import { LocationMap } from "./LocationMap"
import type { LocationData, LocationShare } from "@/lib/location/locationStorage"
import { LocationStorageService } from "@/lib/location/locationStorage"
import { useAuth } from "@/context/AuthContext"
interface LocationViewerProps {
shareToken: string
}
export const LocationViewer: React.FC<LocationViewerProps> = ({ shareToken }) => {
const { fileSystem } = useAuth()
const [location, setLocation] = useState<LocationData | null>(null)
const [share, setShare] = useState<LocationShare | null>(null)
const [loading, setLoading] = useState(true)
const [error, setError] = useState<string | null>(null)
useEffect(() => {
const loadSharedLocation = async () => {
if (!fileSystem) {
setError("File system not available")
setLoading(false)
return
}
try {
const storageService = new LocationStorageService(fileSystem)
await storageService.initialize()
// Get share by token
const shareData = await storageService.getShareByToken(shareToken)
if (!shareData) {
setError("Share not found or expired")
setLoading(false)
return
}
// Check if share is expired
if (shareData.expiresAt && shareData.expiresAt < Date.now()) {
setError("This share has expired")
setLoading(false)
return
}
// Check if max views reached
if (shareData.maxViews && shareData.viewCount >= shareData.maxViews) {
setError("This share has reached its maximum view limit")
setLoading(false)
return
}
// Get location data
const locationData = await storageService.getLocation(shareData.locationId)
if (!locationData) {
setError("Location data not found")
setLoading(false)
return
}
setShare(shareData)
setLocation(locationData)
// Increment view count
await storageService.incrementShareViews(shareData.id)
setLoading(false)
} catch (err) {
console.error("Error loading shared location:", err)
setError("Failed to load shared location")
setLoading(false)
}
}
loadSharedLocation()
}, [shareToken, fileSystem])
if (loading) {
return (
<div className="location-viewer flex items-center justify-center min-h-[400px]">
<div className="flex flex-col items-center gap-3">
<div className="spinner" />
<p className="text-sm text-muted-foreground">Loading shared location...</p>
</div>
</div>
)
}
if (error) {
return (
<div className="location-viewer flex items-center justify-center min-h-[400px]">
<div className="text-center max-w-md">
<div className="text-4xl mb-4">📍</div>
<h2 className="text-xl font-semibold mb-2">Unable to Load Location</h2>
<p className="text-sm text-muted-foreground">{error}</p>
</div>
</div>
)
}
if (!location || !share) {
return null
}
return (
<div className="location-viewer max-w-4xl mx-auto p-6">
<div className="viewer-header mb-6">
<h1 className="text-3xl font-bold text-balance">Shared Location</h1>
<p className="text-sm text-muted-foreground mt-2">Someone has shared their location with you</p>
</div>
<div className="viewer-content space-y-6">
{/* Map Display */}
<LocationMap location={location} precision={share.precision} showAccuracy={true} height="500px" />
{/* Share Info */}
<div className="share-info bg-muted/50 rounded-lg p-4 space-y-2">
<div className="info-row flex justify-between text-sm">
<span className="text-muted-foreground">Precision Level:</span>
<span className="font-medium capitalize">{share.precision}</span>
</div>
<div className="info-row flex justify-between text-sm">
<span className="text-muted-foreground">Views:</span>
<span className="font-medium">
{share.viewCount} {share.maxViews ? `/ ${share.maxViews}` : ""}
</span>
</div>
{share.expiresAt && (
<div className="info-row flex justify-between text-sm">
<span className="text-muted-foreground">Expires:</span>
<span className="font-medium">{new Date(share.expiresAt).toLocaleString()}</span>
</div>
)}
<div className="info-row flex justify-between text-sm">
<span className="text-muted-foreground">Shared:</span>
<span className="font-medium">{new Date(share.createdAt).toLocaleString()}</span>
</div>
</div>
{/* Privacy Notice */}
<div className="privacy-notice bg-primary/5 border border-primary/20 rounded-lg p-4">
<p className="text-xs text-muted-foreground">
This location is shared securely and will expire based on the sender's privacy settings. The location data
is stored in a decentralized filesystem and is only accessible via this unique link.
</p>
</div>
</div>
</div>
)
}

279
src/components/location/ShareLocation.tsx
View File

@ -1,279 +0,0 @@
"use client"
import React, { useState } from "react"
import { LocationCapture } from "./LocationCapture"
import { ShareSettingsComponent } from "./ShareSettings"
import { LocationMap } from "./LocationMap"
import type { LocationData, LocationShare } from "@/lib/location/locationStorage"
import { LocationStorageService, generateShareToken } from "@/lib/location/locationStorage"
import type { ShareSettings } from "@/lib/location/types"
import { useAuth } from "@/context/AuthContext"
export const ShareLocation: React.FC = () => {
const { session, fileSystem } = useAuth()
const [step, setStep] = useState<"capture" | "settings" | "share">("capture")
const [capturedLocation, setCapturedLocation] = useState<LocationData | null>(null)
const [shareSettings, setShareSettings] = useState<ShareSettings>({
duration: 24 * 3600000, // 24 hours
maxViews: null,
precision: "street",
})
const [shareLink, setShareLink] = useState<string | null>(null)
const [isCreatingShare, setIsCreatingShare] = useState(false)
const [error, setError] = useState<string | null>(null)
// Show loading state while auth is initializing
if (session.loading) {
return (
<div className="share-location-loading flex items-center justify-center min-h-[400px]">
<div className="text-center max-w-md">
<div className="text-4xl mb-4 animate-spin">⏳</div>
<h2 className="text-xl font-semibold mb-2">Loading...</h2>
<p className="text-sm text-muted-foreground">Initializing authentication</p>
</div>
</div>
)
}
const handleLocationCaptured = (location: LocationData) => {
setCapturedLocation(location)
setStep("settings")
}
const handleCreateShare = async () => {
if (!capturedLocation || !fileSystem) {
setError("Location or filesystem not available")
return
}
setIsCreatingShare(true)
setError(null)
try {
const storageService = new LocationStorageService(fileSystem)
await storageService.initialize()
// Generate share token
const shareToken = generateShareToken()
// Calculate expiration
const expiresAt = shareSettings.duration ? Date.now() + shareSettings.duration : null
// Update location with expiration
const updatedLocation: LocationData = {
...capturedLocation,
expiresAt,
precision: shareSettings.precision,
}
await storageService.saveLocation(updatedLocation)
// Create share
const share: LocationShare = {
id: crypto.randomUUID(),
locationId: capturedLocation.id,
shareToken,
createdAt: Date.now(),
expiresAt,
maxViews: shareSettings.maxViews,
viewCount: 0,
precision: shareSettings.precision,
}
await storageService.createShare(share)
// Generate share link
const baseUrl = window.location.origin
const link = `${baseUrl}/location/${shareToken}`
setShareLink(link)
setStep("share")
} catch (err) {
console.error("Error creating share:", err)
setError("Failed to create share link")
} finally {
setIsCreatingShare(false)
}
}
const handleCopyLink = async () => {
if (!shareLink) return
try {
await navigator.clipboard.writeText(shareLink)
// Could add a toast notification here
alert("Link copied to clipboard!")
} catch (err) {
console.error("Failed to copy link:", err)
alert("Failed to copy link. Please copy manually.")
}
}
const handleReset = () => {
setStep("capture")
setCapturedLocation(null)
setShareLink(null)
setError(null)
}
if (!session.authed) {
return (
<div className="share-location-auth flex items-center justify-center min-h-[400px]">
<div className="text-center max-w-md">
<div className="text-4xl mb-4">🔒</div>
<h2 className="text-xl font-semibold mb-2">Authentication Required</h2>
<p className="text-sm text-muted-foreground">Please log in to share your location securely</p>
</div>
</div>
)
}
return (
<div className="share-location max-w-4xl mx-auto p-6">
{/* Progress Steps */}
<div className="progress-steps flex items-center justify-center gap-4 mb-8">
{["capture", "settings", "share"].map((s, index) => (
<React.Fragment key={s}>
<div className="step-item flex items-center gap-2">
<div
className={`step-number w-8 h-8 rounded-full flex items-center justify-center text-sm font-medium transition-colors ${
step === s
? "bg-primary text-primary-foreground"
: index < ["capture", "settings", "share"].indexOf(step)
? "bg-primary/20 text-primary"
: "bg-muted text-muted-foreground"
}`}
>
{index + 1}
</div>
<span
className={`step-label text-sm font-medium capitalize ${
step === s ? "text-foreground" : "text-muted-foreground"
}`}
>
{s}
</span>
</div>
{index < 2 && (
<div
className={`step-connector h-0.5 w-12 ${
index < ["capture", "settings", "share"].indexOf(step) ? "bg-primary" : "bg-muted"
}`}
/>
)}
</React.Fragment>
))}
</div>
{/* Error Display */}
{error && (
<div className="error-message bg-destructive/10 border border-destructive/20 rounded-lg p-4 mb-6">
<p className="text-sm text-destructive">{error}</p>
</div>
)}
{/* Step Content */}
<div className="step-content">
{step === "capture" && <LocationCapture onLocationCaptured={handleLocationCaptured} onError={setError} />}
{step === "settings" && capturedLocation && (
<div className="settings-step space-y-6">
<div className="location-preview">
<h3 className="text-lg font-semibold mb-4">Preview Your Location</h3>
<LocationMap
location={capturedLocation}
precision={shareSettings.precision}
showAccuracy={true}
height="300px"
/>
</div>
<ShareSettingsComponent onSettingsChange={setShareSettings} initialSettings={shareSettings} />
<div className="settings-actions flex gap-3">
<button
onClick={() => setStep("capture")}
className="flex-1 px-6 py-3 rounded-lg border border-border hover:bg-muted transition-colors"
>
Back
</button>
<button
onClick={handleCreateShare}
disabled={isCreatingShare}
className="flex-1 px-6 py-3 rounded-lg bg-primary text-primary-foreground hover:bg-primary/90 disabled:opacity-50 disabled:cursor-not-allowed transition-colors font-medium"
>
{isCreatingShare ? "Creating Share..." : "Create Share Link"}
</button>
</div>
</div>
)}
{step === "share" && shareLink && capturedLocation && (
<div className="share-step space-y-6">
<div className="share-success text-center mb-6">
<div className="text-5xl mb-4">✓</div>
<h2 className="text-2xl font-bold mb-2">Share Link Created!</h2>
<p className="text-sm text-muted-foreground">Your location is ready to share securely</p>
</div>
<div className="share-link-box bg-muted/50 rounded-lg p-4 border border-border">
<label className="block text-sm font-medium mb-2">Share Link</label>
<div className="flex gap-2">
<input
type="text"
value={shareLink}
readOnly
className="flex-1 px-3 py-2 rounded-lg border border-border bg-background text-sm"
onClick={(e) => e.currentTarget.select()}
/>
<button
onClick={handleCopyLink}
className="px-4 py-2 rounded-lg bg-primary text-primary-foreground hover:bg-primary/90 transition-colors text-sm font-medium whitespace-nowrap"
>
Copy Link
</button>
</div>
</div>
<div className="share-preview">
<h3 className="text-lg font-semibold mb-4">Location Preview</h3>
<LocationMap
location={capturedLocation}
precision={shareSettings.precision}
showAccuracy={true}
height="300px"
/>
</div>
<div className="share-details bg-muted/50 rounded-lg p-4 space-y-2">
<h4 className="font-medium mb-3">Share Settings</h4>
<div className="detail-row flex justify-between text-sm">
<span className="text-muted-foreground">Precision:</span>
<span className="font-medium capitalize">{shareSettings.precision}</span>
</div>
<div className="detail-row flex justify-between text-sm">
<span className="text-muted-foreground">Duration:</span>
<span className="font-medium">
{shareSettings.duration ? `${shareSettings.duration / 3600000} hours` : "No expiration"}
</span>
</div>
<div className="detail-row flex justify-between text-sm">
<span className="text-muted-foreground">Max Views:</span>
<span className="font-medium">{shareSettings.maxViews || "Unlimited"}</span>
</div>
</div>
<button
onClick={handleReset}
className="w-full px-6 py-3 rounded-lg border border-border hover:bg-muted transition-colors"
>
Share Another Location
</button>
</div>
)}
</div>
</div>
)
}

Some files were not shown because too many files have changed in this diff Show More