Your server. Your data. Free forever.
Unlimited captures, sessions, webhooks, storage. No limits, no cost, no compromise.
Quick start (5 minutes)
Install
Auto-installs Docker if needed, generates secrets, asks for domain and API key.
Start
Starts PostgreSQL + pgvector, Go API, Next.js frontend. Runs migrations automatically.
Use
Navigate to http://your-server:3300 and create your admin account.
Full guide: Docker Compose Deploy
What stays on your server
Everything stays local
- Vault items and metadata
- Hub pages and structure
- Space conversations and artifacts
- Webhook data
- Rules and schemas
- Embeddings and vectors
Only AI prompts leave (if using cloud AI)
- Classification prompts
- Hub materialization prompts
- Space chat messages
Use local Ollama for fully offline AI. Or run without AI—rules-only classification still works.
AI options
Cloud AI
Recommended
Paste your Anthropic API key. No extra containers. Best quality.
Local AI
Fully offline
Ollama with local models. Requires GPU for reasonable speed. Complete privacy.
No AI
Rules only
Classification via rules engine. Hubs are manual. Spaces disabled. Still useful.
Requirements
Minimum
- 2 CPU cores
- 2 GB RAM
- 10 GB storage
- Docker 24.0+
Recommended
- 4 CPU cores
- 4 GB RAM
- 50 GB+ storage
- GPU (for local AI)
Runs on: Linux VPS, home server, Raspberry Pi 4+, Synology NAS, macOS, Windows (WSL2).