Build a public-data source catalog
An organized catalog of relevant datasets with access, freshness, and licensing notes.
A little inspiration is on its way…
For a standing research question: the dot watches a data folder, reruns the notebook on each new file, flags anomalies and sends only the figures that moved.
Community-reported results. Success rates appear after 10 approved reports. Copies are deduplicated per visitor each day. Starter workflows have not been independently tested.
New data arrives every few days and the figures in the draft go stale. OpenAI describes a scientist's dot that reruns the analysis and updates the paper; here is the version you can start with.
Source / inspiration: View credited source
Help the next person know what to expect.
Sign in to share your resultA question, a useful tweak, a little discovery. Start the conversation.
Sign in to join inAn organized catalog of relevant datasets with access, freshness, and licensing notes.
Paste this into a conversation with your dot and replace anything in [square brackets] with your own details. It is a standing job, so say when it should run.
Watch [folder] for new data files matching [pattern]. When one lands, rerun [notebook or script] on your computer with the new data, compare each figure and summary table with the previous run, and investigate any result that moved more than [threshold]. Save the new outputs in [folder]/runs/[date]. Message me only the figures and tables that changed, each with one sentence on the likely cause. Do not edit the manuscript.
Editing the manuscript and sharing outputs outside the folder on ask before taking action. Open the dot's profile, then Customize and Custom rules, describe the action and pick the behaviour. Sensitive actions such as changing a password always stay with you.
The dot's cloud computer for running code, a folder plugin, and memory of the previous run.
Check the environment the dot runs in matches yours (package versions, random seeds). A changed figure can be a changed dependency. On personal plans, review your data-use settings before connecting research data. Dots can make mistakes, including when following your rules; review consequential work before relying on it.
OpenAI's dots announcement and OpenAI's getting-started guide, September 29, 2026. Sundog is an independent community and is not affiliated with OpenAI; this workflow is an untested starting point.
A clear assessment of what a claim says and what its cited evidence supports.