The prompt
Adapt this outline. High risk if mis-exposed. Start read-only.
Help me run a nested-agent homelab on the Bot computer.
1. Inventory processes, ports, entrypoints (names/paths only).
2. Propose least-privilege wiring for the pieces I name.
3. Never expose a port or tunnel URL without my OK.
4. Write ARCHITECTURE.md with restart/kill procedures.
Source: https://x.com/TheVeller/status/2096240503658283399
Rules: no secrets in chat; fail closed on expose. How it's set up
@TheVeller — https://x.com/TheVeller/status/2096240503658283399
High risk if mis-exposed. Start read-only.
- Read-only inventory first.
- Expose nothing until ARCHITECTURE.md exists.
- Distinct from jev-chrome-computer-use on this site.
Why it works
Nested agents are powerful and easy to leak. Inventory plus explicit expose gates keep the lab from becoming an open proxy.
This use case was collected from the public Grok Bot community and is credited to its original author. Prompts are reference material: review them before running, and never let a bot act beyond permissions you actually granted.