XTERM Terminal Setup Guide
Each learner gets a private server, a Node-RED lab, and preinstalled Google Gemini CLI. Secure your API key, paste it into the terminal, send a quick chat message, and confirm your AI assistant is connected. Complete the steps on this page, then use the completion button at the top of this activity to unlock the next mission.
- Personal AI server (with WebTerminal with Chat and Voice mode) + persistent storage
- Gemini + Claude CLI preinstalled (AI Cloud Bot mode)
- Personal Node-RED automation stack
- Confirm terminal login from messaging drawer (
admincredentials) - Create a Gemini API key on aistudio.google.com and paste it into terminal
- Order Gemini to build the starter Node-RED flow
Need WebTerminal (VNC + XTerm) or Node-RED credentials again? Click the speech-bubble icon in the top-right. Toggle messaging drawer — the admin message includes all URLs and passwords. Reply there anytime if you need help.

Open messages, log in as admin, confirm Gemini chat is ready
Create key in AI Studio, paste it into terminal, then press Enter (keyboard or terminal Enter button).
Send a quick greeting (for example, test) and confirm Gemini replies
Use the prepared prompt from Step 4: copy it, paste into Gemini chat, then press Enter (keyboard or terminal Enter button).
When everything below is complete, use the completion button at the top of this activity so the system unlocks the next course level and assignment.
Chat-first interface: write plain English, and Gemini handles execution on your server.
Gemini will tell you when the flow is ready. After that, click Reload and verify the canvas update.
Step 1 · Confirm terminal access first
Start in the messaging drawer. Before creating any API key, make sure the terminal is available and you can sign in successfully.
- Open the messaging drawer (speech bubble) and wait for two admin messages: Node-RED credentials, then WebTerminal is ready.
- After WebTerminal is ready, copy terminal credentials (
admin+ password) from messages. - Refresh the page (F5), sign in to terminal, and save credentials in your browser for next sessions.
- If the terminal frame does not load, click ↗ Open in new tab and continue there.
- When Gemini chat appears and asks for API key activation, move to Step 2.
Step 2 · Generate API key and activate Gemini
Now generate the key in Google AI Studio and paste it into terminal chat so Gemini can start working on your server.
- Open Google AI Studio → API Keys.
- Accept Google terms/licenses if prompted.
- Select an existing project or create a new one.
- Click Create API key, then copy the generated key.
- Return to terminal and paste the key using right-click menu or
Ctrl+Shift+V. - Press Enter and wait for confirmation (for example,
Gemini API key saved).
Step 3 · Verify chat and handle quota limits
Gemini runs in chat mode. Talk to it the same way you talk to mentors — no command syntax required.
If you see
API Error: You have exhausted your daily quota on this model. 
then open terminal settings with /settings and make sure Preview Features = ON.


Next, open the model selector with /model.

Choose Manual mode and select gemini-3-flash-preview. Enable "Remember model for future sessions".

Retry your message. If the limit appears again, wait ~5 minutes. If it still fails, you likely reached the daily limit (RPD) and may need to wait up to 24 hours.

You can also switch to a different Google Cloud project/API key, or ask for help here: https://t.me/sell_systems
If needed, set up billing to unlock higher rate limits.

If you still need help, contact an admin here: https://t.me/sell_systems
Alternative recovery path: use a new project or a new API key.
| Issue | Fix |
|---|---|
| “Invalid API key” | Delete the key in AI Studio → create a new one → paste again. |
| Terminal not responding | Hit Reload, wait 10 seconds, repeat the message. |
| Network hiccup | Restore Wi-Fi/VPN, refresh the page, retry. |
Step 4 · Launch the Starter Flow
Now use the prepared mission prompt below: copy it, paste it into Gemini chat, press Enter (keyboard or terminal Enter button), and supervise execution. This is your first small-win automation.
Paste this into the terminal, then press Enter (or say it in your own words):
Need to inspect the guide manually? Open the starter README in a new tab and follow the same steps.
✦ Gemini reports progress while it works. Answer its questions, confirm prompts, and when it says the flow is ready, click Reload to review changes in Node-RED. If anything is unclear, send a follow-up message or contact an admin — you stay in full control.
Expected output should look like this: 
Gemini usually responds like this when everything is complete:
When GitHub templates expand, we’ll add more advanced prompts here.
Gemini may ask permission to run commands (e.g., ls ). Approve these prompts so Gemini can inspect the environment. You remain the operator and confirm each action.
Use the messaging drawer (top-right speech bubble) to review admin instructions or ask for help. Everything—WebTerminal credentials, Node-RED URL, extra hints—lives there. Or ask here https://t.me/sell_systems
Ready for the next part? Use the completion button at the top of this activity.
