Claude Code auto-memory injects context into every tool call; dev says disable it
nateberkopec · x · 2026-11-16
Developer Nate Berkopec slammed Claude Code's auto-memory feature: screenshots show user memories (e.g. 'prefers Japanese-made goods', 'lives in Japan') and account email being injected as system-reminders into every single tool call, even trivial bash commands. He recommends everyone set CLAUDECODEDISABLEAUTOMEMORY=1. The post sparked a debate with Sentry founder David Cramer, who argued the memory abstraction works well universally — his agent product Junior extracts knowledge from tool responses, and he believes memory shouldn't just live in agents.md or skill files.
More from coding & agent
- Sentry CEO: agent 'memory' is really just knowledge graphs — zeeg · 2026-11-16
- Sentry founder David Cramer defends agent memory: Junior extracts knowledge from tool responses — zeeg · 2026-11-16
- Hooking GSC + GA into an SEO agent: automated search-data diagnosis in practice — yangyi · 2026-11-16
- Neuroscientist pits Fable 5.2, Astra, Hydrogen on REML: wrong code still slips through — pshrink · 2026-11-16
- Testing AI-written parsers with utm_campaign=C%2B%2B: will it refactor C++ down to C? — HaktanSuren · 2026-11-16
- Profound CTO Dylan Babbs on shipping fast and good: two weeks down to two hours — soleio · 2026-11-16