Drowning in scattered user feedback? Have your Dot gather it into a ChatGPT Space tracker and hand bugs to Codex

#What changed

On Oct 23, 2026 the official ChatGPT account on X showed how Rennie Song, an engineer at OpenAI, uses her own Dot to improve Dots: it gathers feedback from social media and user reports, keeps a live tracker in a ChatGPT Space up to date, and she then investigates and fixes the bugs with Codex. In a reply she said the tracker updates automatically every morning at 8am. Source: ChatGPT on X, Rennie Song on X.

#Who is affected

Small product teams and solo builders who get bug reports in many places, such as X replies, support email, and in-app reports, and lose time stitching them together by hand.

#What to do now

1. Give your Dot one standing job: collect feedback about your product from the places you already receive it.
2. Have it write everything into one tracker page in a ChatGPT Space, one row per issue, with a link back to each report.
3. Run it on a daily schedule (Rennie's refreshes at 8am) so the tracker is current when you start work.
4. Pick the top issue and hand it to Codex to investigate and propose a fix, then review the change yourself before it ships.

#What is not confirmed

OpenAI hasn't published step-by-step setup for this workflow; the post shows it only in a short video. Which feedback sources your Dot can read depends on the connectors you have enabled. DotsFeed hasn't tested this loop itself.