Six AI agents ran a company for months — until a laptop lid slept it for 26 hours
Icy_Comfort_6220 · reddit · 2026-10-28
Part 4 of a cost diary: six AI agents run a small publishing company on open-source Paperclip, all on the author's laptop. A post-mortem of failed CEO-agent runs found stale sessions, Wi-Fi drops, a $0.31 timeout, and the laptop lid causing a 26-hour gap between runs. They migrated to Railway before breakfast and hit four undocumented pitfalls: shell eating $ in text, silent empty exports, Windows 260-char path limits (robocopy works, Copy-Item doesn't), and secrets being silently renamed. Day 3 in cloud: costs spiked to $10/day (normal <$2) as agents cleared backlogs, and 16/16 "review productivity" runs still failed due to an unsupported config on a cheap model profile. Lesson: scheduled agents need a machine without a lid — the plumbing, not the model, was the weak point.
More from coding & agent
- Row-Bot: a local 27B model runs a shop's morning ops entirely on one PC via MCP — Acceptable-Object390 · 2026-10-28
- Offboarding AI agents: revoke sessions first, deleting the container isn't enough — ComprehensiveShake76 · 2026-10-28
- The compounding error problem: 95% per-step accuracy means 36% over 20 steps — KelvenSevero0 · 2026-10-28
- Dev shares AI game project's review list: laser bugs, podracer handling fixes queued — jasonkneen · 2026-10-28
- Dev builds online multiplayer castle-storming game with GPT 6.1 Terra, playable on mobile — chongdashu · 2026-10-28
- Self-distillation explained: how a model teaches itself with richer token-level training signals — arpit_bhayani · 2026-10-28