Auto memory on 21 projects: 17 empty, and 3 repos learned the same fix separately
Claude Code's auto memory is on by default and writes to a per-repository directory you probably have never opened. On one machine with 21 such directories, 17 were empty. The other four held two to four notes each, and three of those four had learned the same correction independently, because memory does not cross repositories. We turned the feature off in the one repo where an autonomous agent…
On a single Mac running Claude Code across multiple projects, 17 out of 21 auto memory directories were empty. Three of the remaining four directories contained two to four notes each, with three of those having independently learned the same correction. Claude's auto memory is on by default and writes notes to per-repository directories, which were found on a Mac where the feature was on for 17 projects but found no useful information for most of them.
The other four directories each contained notes about opening URLs in specific Chrome profiles, Terraform version pinning, report dates including the year, and reproducing a failing dependency bot locally. The notes were stored in a user-level CLAUDE.md file, which Claude skips if already present, preventing duplicate learning of the same information in different projects.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
