Solo e-commerce owner turns ChatGPT Scheduled Task into the 'brain' of his five-agent business monitor to dodge Codex quota
coribruebek1212 · reddit · 2026-09-10
An e-commerce owner built a server-side business monitor that pulls analytics, ads, and orders on a schedule, but reasoning calls through his Plus plan's Codex allowance cap out at 20k tokens, killing investigations after 3-4 follow-ups.
The plan: a Scheduled Task wakes on a timer, reads the box via a connector, reasons, and writes back — the server stays the system of record while ChatGPT becomes the thinking part. This also merges five siloed agents (marketing/SEO/finance/support/wholesale): one shared queue could connect a margin drop, rising spend, and support complaints on the same product — structurally impossible for five separate agents.
Tested: an unattended task made two dependent read-only calls via the connector with no approval prompt (a later failure was his own Linux permission misconfig).
Open questions: which model actually runs Scheduled Tasks; whether one run survives 20+ tool calls with reasoning; whether approval prompts can pause unattended runs; and whether anyone runs something similar in production.
More from coding & agent
- Spring AI 2.1.0-RC1 adds tool-call confirmation, pausing agent loops before real-data access — therealdanvega · 2026-10-26
- Morgan Stanley and US banks block third-party AI agents over credentials — thione · 2026-10-26
- METR red-teams OpenAI's agent monitoring, finds bypasses and gaps — thione · 2026-10-26
- OpenAI: coding agents weakened safety monitoring, one lied about it — thione · 2026-10-26
- ConvFill: A 0.6B Talker Buys Time While a Large Reasoner Whispers the Real Answer — andimarafioti · 2026-10-26
- 250 Builders Hack With Codex at Berlin Community Hackathon Ahead of OpenAI DevDay — paw_lean · 2026-10-26