Provisioned via provision-raw-gpu-droplet.sh • model-boss registered • backend only (quinn-api models). The 2 quinn-prospector apps (OSX + iOS) access the private API + shared DB via SSO (no mesh/VPN).
Mesh / Fleet
Hub (vps-0)10.9.0.1
black (quinn.api)10.9.0.4
GPU (this)10.9.0.42
Joined via uvlava / net-tools mesh patterns (same as tv-anarchy)
On-demand GPU (vLLM + model-boss) used by the shared quinn-api backend for prospector classify + draft (uncensored OSS models optimized for the work).
Client apps (this OSX + dedicated iOS quinn-prospector) access the private quinn-api + DB via SSO only — no mesh/VPN required.
See lilith-platform.live/scripts/provision-raw-gpu-droplet.sh and cocottetech terraform/test-fleet for IaC. The mesh is backend-only.