Licensing & Purchase
How do I receive my license key after purchasing?
When you purchase the Hardened Bootstrap or the Sentinel Audit Toolkit, our Merchant Partner automatically registers your secure email address and forwards a digitally signed license file along with Git access instructions within 60 seconds of checkout.
Deployments
Troubleshooting a login redirect loop on VPS
Login redirect loops on public VPS deployments are almost always caused by an HTTP/HTTPS mismatch in the Wiki.js `Site URL` parameter. Ensure the `Site URL` in settings is explicitly configured as `https://wiki.sentinelintegrations.com` and that your reverse proxy is forwarding `X-Forwarded-Proto: https`.
Database
Why PostgreSQL instead of SQLite for public systems?
Our system architecture standard mandates PostgreSQL for public-facing deployments. SQLite locks the entire database file during concurrent writes, resulting in rapid thread exhaustion and system crashes under high traffic loads. PostgreSQL handles parallel transactions natively.
Security Architecture
How is my local lab protected from public servers?
Sentinel Integrations enforces a absolute air-gapped inbound policy. The public VPS on DigitalOcean has zero routing capabilities or VPN connections to your home systems. Backups are pulled securely from the inside out using secure cron triggers on your private systems.
Product Documentation: Hardened Bootstrap
Installing and verifying the Hardened Debian Bootstrap
To deploy, unpack the distribution zip and run the bootstrap script: sudo bash ./bootstrap/install.sh. This hardens your system limits, secures UFW rules, and configures a rootless Docker daemon. To verify compliance, execute the diagnostics utility: python3 ./skills/diagnostics.py --verify-cis.
Product Documentation: Audit Toolkit
Running filesystem watchers and cryptographic Isnad logging
After license validation, enable the background auditing daemon: sudo systemctl enable --now sentinel-audit. This starts the real-time filesystem file-write watcher in /home/sentinel/src/ and begins cryptographically signing all prompt outputs using your local RSA private key.
Product Documentation: Small Business CRM
Initializing PostgreSQL schemas and Docker Compose pipelines
Launch the local containerized CRM database by executing: docker compose up -d postgres inside the /docker/ folder. This bootstraps the PostgreSQL daemon. Then, initialize tables by running psql -h localhost -U n8n_user -f ./schema/init_crm.sql. Lead pipelines parse sales in real-time.
Product Documentation: Video Production
Auto-compiling video voiceovers and FFmpeg storyboards
Place your storyboard table inside /templates/storyboard.md. Run the python orchestrator: python3 ./scripts/pipeline.py --input /templates/storyboard.md --voice kokoro. The script polls your local Kokoro speech engine to synthesize voice narration clips and uses FFmpeg to compile a unified, master audio track automatically.
Product Documentation: Sentinel Mesh
Configuring Tailscale overlays and secure token delegation
Define your mesh nodes inside /vpn/nodes.yaml. Apply the secure VPN overlay configuration: sudo tailscale up --advertise-routes=10.1.1.0/24. For cross-node execution, the Paperclip controller coordinates WebSocket connections and elevates administrative privileges via non-interactive sudoers.
Product Documentation: RAG Pro
Setting up PGVector database and handling model rate limits
Initialize PGVector or Qdrant inside /vector-db/ via docker compose. All model queries should be routed through the python API handler: from api_handler import SecureModelClient. This client automatically intercepts rate-limiting errors (HTTP 429) and performs exponential backoffs.
Product Documentation: Enterprise Node (SEN)
Staging the 5-layer sovereign AI enterprise stack
Run docker compose up -d in /opt/sentinel-enterprise-node/ to deploy the unified stack. This initializes: (1) Debian host hardening, (2) Tailscale networking overlay, (3) NATS message broker for agentic pub-sub, (4) Infisical local secrets vault, and (5) Ollama container for private on-premise inference.