model-catalog
Use when working with 3D model files (STL), print tracking, slicer profiles, or managing the model-catalog homelab service at model-catalog.internal.lakehouse.wtf
When & Why to Use This Skill
The Model Catalog skill provides a comprehensive interface for managing a 3D printing homelab environment and digital asset library. It enables users to interact with a FastAPI-based service to organize STL files, track print job history, and manage slicer profiles. By bridging the gap between 3D modeling and physical production, this skill streamlines the workflow for makers and developers, offering tools for model versioning, print success analysis, and remote service management via REST API and CLI integrations.
Use Cases
- Digital Asset Management: Uploading, tagging, and searching for 3D model files (STL) to maintain an organized, searchable homelab catalog.
- Print Job Lifecycle Tracking: Logging print attempts, success rates, and failure reasons to optimize printer settings and material usage.
- Slicer Profile Centralization: Managing and downloading PrusaSlicer .ini profiles to ensure consistent print configurations across different workstations.
- Homelab DevOps & Maintenance: Monitoring service health, viewing logs, and managing the deployment of the model-catalog service on LXC containers and Proxmox environments.
- Automated Print Monitoring: Integrating with PrusaLink to query real-time printer status and aggregate print statistics for performance analysis.
| name | model-catalog |
|---|---|
| description | Use when working with 3D model files (STL), print tracking, slicer profiles, or managing the model-catalog homelab service at model-catalog.internal.lakehouse.wtf |
Model Catalog
Overview
A homelab FastAPI service for cataloging 3D models, tracking print history, managing slicer profiles, and storing print photos. Deployed as LXC 210 with Vue.js frontend and REST API.
When to Use
Use this skill when:
- Uploading or searching for 3D model files (.stl)
- Tracking print jobs, success rates, or failures
- Managing PrusaSlicer profiles
- Querying print statistics
- Working with the model-catalog service deployment
- Integrating with PrusaLink/Prusa Connect
Quick Reference
Service Information
| Component | Details |
|---|---|
| URL | https://model-catalog.internal.lakehouse.wtf |
| Container | LXC 210 on proxmox3 (192.168.1.211:8000) |
| Frontend | Vue 3 + TypeScript (served at /) |
| API Docs | /docs (Swagger UI) |
| Database | SQLite at /var/lib/model-catalog/catalog.db |
| Storage | NFS /mnt/models from TrueNAS |
API Endpoints
| Endpoint | Method | Purpose |
|---|---|---|
/api/health |
GET | Health check |
/api/models |
GET, POST | List/upload models |
/api/models/{id} |
GET, DELETE | Get/delete model |
/api/models/{id}/file |
GET | Download STL file |
/api/prints |
GET, POST | List/create print jobs |
/api/prints/{id} |
GET, PATCH | Get/update print |
/api/prints/{id}/complete |
POST | Mark print successful |
/api/prints/{id}/cancel |
POST | Cancel print |
/api/profiles |
GET, POST | List/create slicer profiles |
/api/profiles/{id}/ini |
GET | Download profile INI |
/api/tags |
GET, POST | List/create tags |
/api/stats |
GET | Overall statistics |
/api/printer/status |
GET | Printer status (via PrusaLink) |
CLI Commands
The mc CLI is available in the container:
# List models
mc list
mc list --tag "functional-prints"
# Search models
mc search "benchy"
# Add model
mc add /path/to/model.stl --name "Test Print" --tags "test,calibration"
# Show details
mc show <model-id>
# Export model
mc export <model-id> /path/to/output.stl
# Delete model
mc delete <model-id>
# Tag operations
mc tag add <model-id> "new-tag"
mc tag remove <model-id> "old-tag"
Common Workflows
Uploading a Model
Via API:
curl -X POST https://model-catalog.internal.lakehouse.wtf/api/models \
-F "file=@model.stl" \
-F "name=My Model" \
-F "description=Test print" \
-F "tags=functional,test"
Via CLI (on container):
ssh root@192.168.1.211
mc add /mnt/models/myfile.stl --name "My Model" --tags "functional,test"
Tracking a Print Job
# Start print
curl -X POST https://model-catalog.internal.lakehouse.wtf/api/prints \
-H "Content-Type: application/json" \
-d '{
"model_id": "uuid-here",
"profile_id": "profile-uuid",
"notes": "First test print"
}'
# Complete print (success)
curl -X POST https://model-catalog.internal.lakehouse.wtf/api/prints/{print-id}/complete \
-H "Content-Type: application/json" \
-d '{"result": "success", "notes": "Perfect print"}'
# Mark failed
curl -X POST https://model-catalog.internal.lakehouse.wtf/api/prints/{print-id}/complete \
-H "Content-Type: application/json" \
-d '{"result": "failure", "notes": "Layer shifting at 50%"}'
Managing Slicer Profiles
# Import PrusaSlicer profile
curl -X POST https://model-catalog.internal.lakehouse.wtf/api/profiles/import \
-F "file=@profile.ini" \
-F "name=PLA Standard"
# List profiles
curl https://model-catalog.internal.lakehouse.wtf/api/profiles
# Download profile INI
curl https://model-catalog.internal.lakehouse.wtf/api/profiles/{id}/ini -o profile.ini
File Locations (Production)
LXC 210 (192.168.1.211):
/var/lib/model-catalog/catalog.db # SQLite database
/opt/model-catalog/ # Application code
/etc/model-catalog/env # Configuration
/mnt/models/ # NFS mount (TrueNAS)
├── models/{id}/original.stl.zst # Compressed STL files
├── gcode/{id}.gcode.zst # Generated G-code (temp)
├── photos/{id}/finished.jpg # Print photos
└── profiles/{id}.ini # Slicer profiles
TrueNAS:
/mnt/truenas_nvme/active_projects/model-catalog/ # NFS backing store
Service Management
Check service status:
ssh root@192.168.1.126 "pct exec 210 -- systemctl status model-catalog"
Restart service:
ssh root@192.168.1.126 "pct exec 210 -- systemctl restart model-catalog"
View logs:
ssh root@192.168.1.126 "pct exec 210 -- journalctl -u model-catalog -f"
Update deployment:
# Package updated code
cd /home/dev/workspace/model-catalog
tar czf /tmp/update.tar.gz --exclude='.git' app/ frontend/dist/
# Deploy to container
scp /tmp/update.tar.gz root@192.168.1.126:/tmp/
ssh root@192.168.1.126 "
pct push 210 /tmp/update.tar.gz /tmp/update.tar.gz
pct exec 210 -- bash -c 'cd /opt/model-catalog && tar xzf /tmp/update.tar.gz'
pct exec 210 -- chown -R model-catalog:model-catalog /opt/model-catalog
pct exec 210 -- systemctl restart model-catalog
"
Architecture
┌─────────────────────────────────────────────────────────────┐
│ Browser │
│ https://model-catalog.internal.lakehouse.wtf │
└──────────────────┬──────────────────────────────────────────┘
│ DNS → 192.168.1.101 (Traefik VIP)
▼
┌─────────────────────────────────────────────────────────────┐
│ Traefik HA (VIP) │
│ Primary: 192.168.1.110 | Secondary: 192.168.1.103 │
└──────────────────┬──────────────────────────────────────────┘
│ Proxy → 192.168.1.211:8000
▼
┌─────────────────────────────────────────────────────────────┐
│ LXC 210 (model-catalog) - 192.168.1.211 │
│ │
│ ┌────────────────┐ ┌──────────────────┐ │
│ │ FastAPI (8000) │ │ Vue.js Frontend │ │
│ │ - API │ │ - / │ │
│ │ - /api/* │ │ - /assets/* │ │
│ └────────┬───────┘ └──────────────────┘ │
│ │ │
│ ▼ │
│ ┌────────────────────────────────────┐ │
│ │ SQLite (/var/lib/model-catalog/) │ │
│ └────────────────────────────────────┘ │
└─────────────────┬───────────────────────────────────────────┘
│ NFS mount → /mnt/models
▼
┌─────────────────────────────────────────────────────────────┐
│ TrueNAS │
│ /mnt/truenas_nvme/active_projects/model-catalog/ │
│ - STL files (zstd compressed) │
│ - G-code (temporary, deleted after success) │
│ - Print photos (JPG from PrusaConnect) │
│ - Slicer profiles (INI) │
└─────────────────────────────────────────────────────────────┘
Environment Variables
Set in /etc/model-catalog/env:
| Variable | Default | Description |
|---|---|---|
MC_DATABASE_PATH |
/var/lib/model-catalog/catalog.db |
SQLite database path |
MC_STORAGE_PATH |
/mnt/models |
NFS mount point for files |
MC_COMPRESSION_LEVEL |
3 |
Zstd compression (1-22) |
MC_DEBUG |
false |
Enable debug logging |
MC_PRUSASLICER_PATH |
/usr/bin/prusa-slicer |
PrusaSlicer CLI path |
Integration Examples
From Claude Code
Check if model exists:
curl -s https://model-catalog.internal.lakehouse.wtf/api/models?search=benchy | jq '.items[0]'
Upload from local file:
curl -X POST https://model-catalog.internal.lakehouse.wtf/api/models \
-F "file=@/path/to/model.stl" \
-F "name=$(basename model.stl .stl)" \
-F "tags=claude-generated"
Get print statistics:
curl -s https://model-catalog.internal.lakehouse.wtf/api/stats | jq
From Automation Scripts
Track automated print:
import httpx
async with httpx.AsyncClient() as client:
# Create print job
response = await client.post(
"https://model-catalog.internal.lakehouse.wtf/api/prints",
json={
"model_id": model_uuid,
"profile_id": profile_uuid,
"notes": "Automated nightly print"
}
)
print_id = response.json()["id"]
# Mark complete when done
await client.post(
f"https://model-catalog.internal.lakehouse.wtf/api/prints/{print_id}/complete",
json={"result": "success"}
)
Common Issues
Frontend Returns 404
Problem: Root path returns {"detail":"Not Found"}
Solution: Frontend not configured to serve from /. Check:
# Verify frontend files deployed
ssh root@192.168.1.126 "pct exec 210 -- ls -la /opt/model-catalog/frontend/dist/"
# Check if main.py has static file mounting
ssh root@192.168.1.126 "pct exec 210 -- grep -A 5 'StaticFiles' /opt/model-catalog/app/main.py"
Service Won't Start
Problem: systemctl status model-catalog shows failed
Solution: Check logs:
ssh root@192.168.1.126 "pct exec 210 -- journalctl -u model-catalog -n 50"
Common causes:
- NFS mount not available (
/mnt/modelsmissing) - Database path not writable
- Python dependencies missing
Traefik Returns 503
Problem: Traefik route configured but backend unreachable
Solution:
# Check backend is responding
curl http://192.168.1.211:8000/api/health
# Verify Traefik route points to correct IP
ssh root@192.168.1.110 "grep -A 5 'model-catalog:' /etc/traefik/dynamic/model-catalog.yml"
Development
Local development (in workspace):
cd /home/dev/workspace/model-catalog
# Run backend
uvicorn app.main:app --reload --port 8000
# Run frontend dev server (separate terminal)
cd frontend
npm run dev
# Run tests
pytest
pytest --cov=app
Testing API locally:
# Health check
curl http://localhost:8000/api/health
# Interactive docs
open http://localhost:8000/docs
Related Documentation
- Project:
/home/dev/workspace/model-catalog/ - Design Doc:
/home/dev/workspace/model-catalog/docs/DESIGN.md - Deployment:
/home/dev/workspace/model-catalog/docs/DEPLOYMENT.md - Acceptance Criteria:
/home/dev/workspace/model-catalog/docs/ACCEPTANCE.md - Issue Tracking:
bd list(beads in/home/dev/workspace/model-catalog/.beads/)