brAIngrid node - share your GPU, get paid
=========================================

1. Install LM Studio (lmstudio.ai) and download a model your GPU can hold:
     8 GB VRAM  -> an 8B model at Q4 (Llama 3.1 8B Instruct, Qwen3 8B)
     16-24 GB   -> up to a ~32B model at Q4
   That's all - the brAIngrid node starts LM Studio's server by itself
   whenever it's needed.

2. Get your join code: sign up / sign in at
     https://braingrid.us/dashboard
   The "Node setup" card shows a ready-made node.toml with your join code.
   Replace the node.toml in this folder with that snippet (or just edit the
   enroll_token line to your own gpu-... code).

3. Double-click aipool-tray.exe
   A brAIngrid icon appears in the system tray and sharing starts.
   Right-click it for START SHARING / STOP SHARING / Open dashboard / Quit.

Your GPU appears on your dashboard within seconds and earns for your account.

Safety, in plain words: the agent only dials OUT (no port forwarding, no
inbound ports), and the only thing the network can ask it to do is run an
inference request against your local LM Studio and stream the result back.
There is no remote-command surface. STOP SHARING or Quit ends it instantly;
nothing runs at startup unless you set that up yourself.
