AI agent stories that use Codex

4 stories name Codex among the tools used. By outcome: Get tasks done 3 and Save money 1. Countries: Country unknown 1, Singapore 1, United Kingdom 1 and United States 1. Other tools named in 2 or more of them: Claude Code 2 and Gmail 2. Results are shown as claimed in each public source.

Get tasks doneBusiness size: SoloIncludes prompt

A consultant running two companies hands his tax letters, bank queries and company filings to AI agents and says they saved him at least 2 days in a month

Peter Hartree @peterhartree · Country unknown · Claude Code, Codex, Gmail, WhatsApp, Google Drive

He sends a screenshot, PDF or forwarded message with a short request. The agents have access to his Gmail, WhatsApp and Drive and can use a browser. Examples he lists: tracing an unexplained bank seizure to a tax charge, checking a tax portal and posting a document by registered mail, reviewing and approving year-end company accounts, and flagging terms in a merger engagement letter that contradicted what was agreed by email.

Result, as claimed

He estimates the AIs saved him at least 2 days of work over the past month. He says the error rate seems similar to or lower than his own and that agents rival a human PA for many tasks.

Get tasks doneBusiness size: SoloIncludes prompt

London solo consultant runs his client notes, tasks and webinars from one AI-maintained folder and says proposals now take 30 minutes, not 4 hours

Nick Zervoudis · United Kingdom · Claude Code, Codex, Obsidian, Granola, Wispr Flow, Telegram, Gmail, Circle, Luma, Lovable

Working from one folder of plain text files, Claude Code or Codex updates a note for each company, person and opportunity after every synced call transcript, creates tasks with priority and an estimated duration, starts promised work such as intro emails, and writes a daily brief that also arrives on Telegram. For webinars it makes the banners, drafts the descriptions and publishes the event to Circle and three Luma calendars once he says publish.

Result, as claimed

His own estimates, shared with Create With: proposal prep down from about 4 hours to about 30 minutes, an event banner and descriptions from about 40 minutes to about 5, and delivery effort down about 45% on one engagement. Average monthly revenue in the first half of 2026 ran at 2.4 times the 2025 monthly average, which he credits partly to a young business growing. His community is on course for about 33 virtual events in 2026, against 8 in 2025. The setup costs around £250 a month.

Get tasks doneBusiness size: SoloIncludes prompt

An Iowa builder says his Codex listing pipeline costs about $0.85 a card and paid back its build by the third batch

Matt Murphy · United States · Codex, OpenAI models

Matt Murphy sells trading cards on eBay. He built a pipeline with Codex that wakes up each night and works through a batch of 50 photographed cards: it identifies each card, researches what it sold for and drafts the listing. He reads and publishes every draft himself, and it stops and asks when it isn't sure.

Result, as claimed

By hand each card took him 5 to 10 minutes. The pipeline's median batch run of 50 cards cost $42.39 at API list prices, about $0.85 a card, plus about 30 seconds of his review. Valuing his time at $150 an hour, he puts the saving at $10.40 to $22.90 a card and $520 to $1,145 a batch, so the $1,200 build paid for itself by the third batch.

Save moneyBusiness size: SoloIncludes prompt

A Singaporean with no coding skills used Codex to build a tool that picks the best credit card for miles at 3,300 merchants

Hosan Swee (KiasuMiles) · Singapore · Codex, ChatGPT

Hosan Swee and his wife share about a dozen credit cards and kept losing miles by picking the wrong one. He used Codex and ChatGPT to build KiasuMiles. Codex scanned bank and merchant websites to build a structured database of card rules, including minimum spend, miles per dollar, eligible categories and monthly caps. People connect KiasuMiles to ChatGPT, Claude or Codex and ask which card to use at a given shop. He built it feature by feature with specific prompts and tested the logic on a few cards first.

Result, as claimed

The Straits Times reports the tool tracks around 50 credit cards and over 3,300 merchants in Singapore. Compiling the card rules was done in minutes with Codex, compared with the hours it would take by hand. No miles or dollar savings figure is given.

Other tools