How to use Codex to research mobile plans remotely
Codex researched mobile plans from Jane Arandelovic’s remote Mac and saved documents; household reception and any contract switch stayed human decisions.
linkedin.com
Suggested prompt
Suggested prompt written by Agents In Use, not the source author. Replace bracketed inputs and review tool permissions first.
Compare mobile plans against these approved data, price and contract criteria from my remote Mac. Save a source-linked document with plan terms, rewards conditions and uncertainties. Separate advertised network coverage from unverified reception inside my home, and pause for my review before any login, contract change or purchase.
Reach people exploring AI agents
See placements →Evidence at a glance
Sourced account- Source author
- Jane Arandelovic
- Original account
- Read the source ↗
- Published
- Source link checked
- Reported result
- Codex completed remote mobile-plan research and saved documents; no contract switch or measured saving is reported.
- Human involvement
- The owner defined the criteria and retained coverage, reputation and purchasing decisions; actual reception in her lounge remained outside the agent’s evidence.
A linked account supports attribution. It does not mean we independently reproduced its result.
Report an issue with this case
Open correction emailOpens your email app with the case link included. Nothing is sent automatically.
How the agent did it
- Arandelovic supplied mobile-plan criteria from the phone while Codex operated her remote Mac.
- Codex researched alternatives and saved local comparison documents, including an unfamiliar provider and rewards connection.
- The owner retained review of reputation, real household coverage and whether any plan was worth switching to.
Jane Arandelovic used Codex from her phone to research mobile plans while her Mac remained at home. She describes delegating the task from a parked car, with the remote computer doing the browsing and retaining the output as local documents.
The owner supplied comparison criteria rather than asking for an unqualified cheapest-plan answer. Codex found an unfamiliar provider and surfaced a grocery-rewards connection that might change the effective cost. The article does not publish a completed switch or a reconciled saving.
Arandelovic still needed to assess reviews and coverage. In particular, an agent comparing published plan terms could not establish whether reception would work in a known weak spot in her lounge. That local constraint remained a human judgment, not a fact supplied by a generic comparison table.
The account distinguishes using ordinary ChatGPT to think through a question from Codex taking action on a computer and saving files. The demonstrated result is a remotely executed research task with durable documents. No access to her inbox or calendar was needed for this run, and no new mobile contract or purchase is reported.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: LinkedIn · Jane Arandelovic
Frequently asked questions
Where did the browsing run?
Codex operated the author’s Mac at home while she delegated through her phone.
Did the agent verify reception inside the home?
No. The author identifies lounge coverage as a remaining real-world judgment.
Did she buy a new plan?
The verified account reports research and review, not a completed switch.
More details
About this account
Arandelovic supplied mobile-plan criteria from the phone while Codex operated her remote Mac. Codex researched alternatives and saved local comparison documents, including an unfamiliar provider and rewards connection.
Full primary dated article opened. Actual remote research is distinguished from ordinary ChatGPT discussion. Potential reward-linked cost improvement is not measured savings, and no switch or private account access is inferred. Suggested prompt written by Agents In Use; no verbatim original task prompt was supplied.
Credit: Jane Arandelovic — LinkedIn
Source published
Added to Agents In Use
Source checked
Tools and topics
Codex · ChatGPT mobile app · Remote Mac · Local documents
mobile-plan research · remote delegation · coverage review
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.