Let a laptop control coding agents that run on a home PC

Goal: Let a laptop control coding agents that run on a home PC
Setup: agents used in source: Claude, Codex, Pi, OpenCode, Hermes Agent.
Before you start: A home Windows PC with WSL2 Ubuntu.
Steps:
1. Run Docker inside WSL2 Ubuntu, with limits: Not Docker Desktop; every container gets memory and CPU limits; add a Windows-side keepalive.
2. Connect devices with Tailscale; browser tools via Cloudflare Tunnel + Access: No open router ports.
3. Run the Herdr server on the PC, client on the laptop: Each task gets its own pane, usually in its own git worktree; panes survive closing the lid.
4. Put a gateway (9Router) in front of every model: All agent CLIs talk to one localhost endpoint; combos handle fallback, round robin and fusion.
5. Launch agents against the gateway with 9agent: Sets Claude Code env vars to gateway models; --sandbox runs in Docker with only the working dir mounted.
6. Schedule Hermes jobs for upkeep: Nightly CLI updates, security PR scans every six hours, a 30-minute autoheal. Hermes drafts, you send.
7. Turn corrections into skills: Each fix goes into a skill file with a growing pitfalls section.
Rules: ask me before anything irreversible (sending, posting, paying, deleting); don’t invent data, results or tools I haven’t given you; tell me what you couldn’t verify.
Based on: Nadim Tuhin — https://www.nadimtuhin.com/blog/ai-coding-agent-setup (summarised from source; the source has no explicit step list)

Steps are our summary of the source (it has no explicit step list). Includes a “Based on” credit line.