Pull published-post stats without editing the dashboard

Charlie Hills used Codex Computer Use to read Substack post metrics across every published entry and verify the count against the dashboard total.

Done with Codex· Intermediate

What they did
Charlie set up Codex Computer Use, enabled it for the needed apps, and granted the Mac permissions it required for screen viewing and control. Then, with the Substack dashboard open in Chrome, they asked Codex to scan every page of the published-post list, record the figures, and save a report with the date on it. The instructions also told it to cross-check the number of posts against the dashboard total and to leave posts and subscriber details untouched.
How it went
Codex identified and stored 122 unique posts, then confirmed that figure against the dashboard total. On a second run on 16 September, it went through all five pages again, picked up revised numbers, and still ended at 122 posts. The numbers reflect totals over the life of each post, which means older entries have had more time to build results.
Tips
This works well when you need a read-only scrape from a product without an MCP connector. The figures can help point to patterns, but they do not explain why a post did well.

Try it yourself with Codex

Open my [Substack or other platform] dashboard in [browser], then review every published item and pull these fields for each one: [fields]. Save everything to [file path] as a dated spreadsheet, ordered by [date or metric]. Check the total item count against the dashboard total, and report any mismatch. Before you do anything that would change, submit, delete, send, or pay for something, stop and ask me to approve it. Do not edit any content or open subscriber details.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.