cd /news/ai-agents/local-ai-agents-investigate-an-unkno… · home › topics › ai-agents › article
[ARTICLE · art-174231] src=github.com ↗ pub= topic=ai-agents verified=true sentiment=↑ positive

Local AI agents investigate an unknown machine attached to a serial port

A local LLM agent running on a Raspberry Pi identified an anonymous device wired to its serial port as a Motorola 68000 single-board trainer running "TUTOR 1.3" in roughly 5.5 hours, according to a published write-up and run logs in the explorer-of-serial repository. The solver agent, Qwen3.8-Flash-Next on a DGX Spark, mapped the machine's memory by deliberately triggering bus errors, decoded its monitor command set, and assembled and ran its own program on the 68000 hardware across 87 turns, with a second independent LLM critic on another box rejecting the verdict once for citation gaps before approving at round 2. The agent ran inside a systemd network sandbox (IPAddressDeny=any plus an allowlist) with a netcheck gate that blocks the run if the jail leaks, and later hung the device's CPU while probing an I/O corner after all goals had been met.

recommendation

Run the solver and an independent critic on separate model servers, require evidence chains for every claim, and persist notes to a findings file so long runs survive context compaction.

more in playbook: Designing agent systems
read3 min views2 publishedNov 22, 2026
Local AI agents investigate an unknown machine attached to a serial port
Image: Michielbdejong (auto-discovered)

Local AI agents, with only local LLMs, investigate an unknown machine attached to a serial port. It doesn't have Internet access, does not have access to the device's documentation. There's no human to help. Just two local LLMs, an SSH session into a Raspberry Pi, and that unknown device.

I wired an anonymous device — “X” — to a Raspberry Pi's serial port, told an LLM agent: find out what X is, reverse-engineer its protocol, and run your own program on it, and walked away. The agent ran inside a network jail (no way out except two model servers and the Pi), and every claim it made had to be verified, based on evidence by a second, independent LLM with fresh context on another box.

It worked. In ~5.5 hours, Qwen3.8-Flash-Next (on a DGX Spark) discovered that X was a Motorola 68000 single-board trainer running “TUTOR 1.3”. It first suspected the machine was KIM-1 with a similar ROM, but later corrected the course. It mapped X's memory by deliberately triggering bus errors, decoded its monitor's command set line-by-line, assembled and ran its own program on 68000 hardware, and wrote a verdict the critic accepted — RESULT.md. Then, probing one I/O corner too deep, it hung X's CPU. Which was not a problem at that point, as all the goals had been by then achieved.

Both roles are OpenCode agents, driven headlessly by go.sh via opencode run; OpenCode's permission layer is also the first jail layer.

  • Solver — theexplorer agent: plans, SSHes into the Pi, pokes the serial line, keeps a durable notes file (FINDINGS.md ) that survives context compactions over a 12-hour run.
  • Critic — theverifier subagent on a second model server: adversarial final gate, ≤3 rounds, grades file pointers + evidence chains, vetoes unsupported claims. It rejected and forced fixes along the way.
  • Jail — systemd network sandbox (IPAddressDeny=any + allowlist) with a hardnetcheck gate: if the jail leaks, the run refuses to start. The rules are also in the prompt — and the agent never tried to escape either. Details:docs/JAIL.md .
  • Ground rules —AGENTS.md (protocol) +task.md (mission): read evidence or perish; “I don't know, and here's how to revive it” beats a confident fabrication.

Three full runs are published under runs/ — the headline run first. Each one holds RESULT.md (the verdict), FINDINGS.md (the working notebook) and the raw console logs — full transcripts, nothing summarised, only home-network details mechanically renamed:

  • runs/2026-11-19-68000-trainer/ — the run described above (Qwen3.8-Flash-Next , 87 turns): the task won, the 8-stepevidence/ chain included; the final gate rejected it once on citation gaps and approved at round 2.
  • runs/2026-11-20-68000-deep-dive/ — the same machine re-examined against every byte of a new raw transcript; the critic only signed off at round 3, having caught two rounds of over-claims, and recorded "no new findings" — verification, not rediscovery.
  • runs/2026-11-17-first-contact/ — the very first end-to-end completion: ~39 hours over 8 resumes, and the one that first pinned X as a 68000-family system running TUTOR.
  • The harness: go.sh (run entry),scripts/ (jail, lifecycle, citation auditor, live console watcher),config/ .

Requires the OpenCode CLI, two OpenAI-compatible model servers (e.g. llama.cpp's llama-server), and a Linux box with systemd-networkd if you want the jail. Local weights only — bring your own endpoints:

cp config/explorer.example.json config/explorer.$(id -un).json   # point at your model servers
cp .env.example .env                                              # point at your target
sudo scripts/install-jail.sh                                      # optional: enable the jail
TASK_FILE=task.md ./go.sh
scripts/commentate.py --web --db                                  # narrate it live on a page

Full lifecycle docs: HOW_TO_RUN.md.

No devices were permanently harmed. X just needs someone to pull its plug — the one thing the agent wasn't allowed to do.

MIT licensed. Write-ups and logs included.

── more in #ai-agents 4 stories · sorted by recency
── more on @qwen3.8-flash-next 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/local-ai-agents-inve…] indexed:0 read:3min 2026-11-22 · —